Synthesis, antioxidant, anticholinesterase activities and molecular docking studies of coumaryl 1,3- selenazoles derivatives

Inhibition of acetylcholinesterase (AChE) enzyme is a known procedure to treat severe Alzheimer's disease through increasing the acetylcholine level in the brain and thus slowing down the progression of Alzheimer's symptoms. The approved medications are only considered as palliative and addressed some reported deficiencies. Therefore, the demand for safe and effective compounds is substantially increasing. A newly series of coumaryl 1,3-selenazoles derivatives was synthesized in four steps. Then, their antioxidant activities were evaluated using DPPH, ABTS cation radical scavenging assay and cupric reducing antioxidant capacities (CUPRAC). The anticholinesterase activities were evaluated using the Ellman method. Then, the docking studies were carried out to explain the possible correlation between in vitro anticholinesterase activity results and the ligand-receptor interactions. Ten new coumaryl 1,3-selenazoles (5a-5d series and 6a-6f series) derivatives were successfully synthesized. The DPPH radical scavenging assay showed that all tested compounds have IC50 value > 200 μM, for ABTS cation radical scavenging assay the IC50 value > 1000 μM and for CUPRAC assay the IC50 value > 200 μM. Compound 5c was found to be the most active compound against AChE and BChE in its series with IC50 value for AChE is 99.76 μM and IC50 for BChE is 140.28 μM while 6b exhibited the most potent inhibition in its series with IC50 value for AChE is 56.01 μM and IC50 for BChE is 121.34 μM. Besides, the docking studies showed that compound 5c and 6b formed π-π stacking interaction with aromatic residues at the active site of AChE and BChE, which is responsible for inhibiting the enzymes. This shows that the synthesized compounds contain skeletal structures that can interact and inhibit within the enzymes active site.


Introduction
Alzheimer is categorized as a dementia disease which prominently attacks the elderly aged between 65 years and above [1,2]. Generally, Alzheimer patients will experience loss of memory that is accompanied by cognitive deterioration -a deteriorating in the ability to think and learn which contributed to the difficulty in speaking and walking, henceforth these drawbacks conceivably lead to death [1,3,4]. Among elderly patients attending medical clinics in Universiti Kebangsaan Malaysia Medical Centre, the prevalence of mild cognitive impairment was apparent and is believed to be a precursor to Alzheimer disease [5] Three main factors affect Alzheimer disease, including a decrease of acetylcholine level, formation of β-amyloid plaque and the digression of neurofibrillary in human brain [3,4,6]. In addition, oxidative stress plays a crucial role in the pathogeny of Alzheimer 's disease [7,8]. Nevertheless, consideration of the underlying factors and the progressions of this disease have not been fully understood [9][10][11]. Notably, four acetylcholinesterase (AChE) inhibitors have been commercially used, namely tacrine (Cognex® ), rivastigmine (Exelon® ), donepezil (Aricept® ) and galantamine (Razadyne® ) (Fig 1). Apart from that memantine, N-methyl-ᴅ-aspartate (NMDA) receptor agonist, has also been employed to moderate Alzheimer's disease in mitigating additional neurological conditions by acting as neuroprotective agent that positively impacts both neurodegenerative and vascular processes [2,[12][13][14]. Tacrine was the first AChE inhibitor that has been approved in 1993, but the usage was withdrawn because of highly toxic to the liver and demonstrated low bioavailability [13,14]. In 1996, another AChE inhibitor namely donepezil was approved for Alzheimer's treatment. Donepezil is reversible and highly active with low toxicity. Galantamine -an alkaloid compound found in Galanthus woronaii species and Amaryllidaceae family, is the only AChE inhibitor derived from natural products. This reversible alkaloid is highly competitive and is a selective AChE inhibitor. The treatment with galantamine potentially showed good results, but thus far, itless effective than tacrine.
Due to complex structure, the synthesis of galantamine requires series of prolonged passive reactions [14,15]. Conversely, rivastigmine is a pseudo-irreversible AChE inhibitor with a lower effect towards butyrylcholinesterase enzyme (BChE). It is noticeable that BChE is responsible for the hydrolysis of acetylcholine to choline by lowering the level of acetylcholine in the brain. Therefore, the inhibition of BChE also plays an essential role in the treatment of Alzheimer's disease. Further progress in AChE inhibitor drug discovery can be seen through the development of memantine -the last drug approved that acts through a different mechanism[ [2,14]]. Despite the availability of these drugs, the 3 of 22 remedies remain palliative with side effects, such as nausea, anorexia, vomiting and diarrh oea [16]. Therefore, there is a need to find a new potent AChE inhibitor with low toxicity effects. AChE has two binding sites, namely an anionic site of the catalyst and the peripheral anionic site that is linked by the spacer group. AChE inhibitors bind through one or both sites with the latter leading to the dual inhibitions of AChE. Thus, it is a considerable prospect to synthesize a compound that can interact with both sites for dual inhibition by linking two different moieties through an appropriate spacer.
The coumarin ring is a heterocyclic moiety that previously demonstrated to be a potential anti-AChE and favourable by the optimal synthetic accessibility [17][18][19].Coumarin derivatives are also among various synthetic compounds that have been tested for AChE inhibitor [17,20]. The coumarin pharmacophore has been predicted as one of the moieties as this heterocyclic compound can form parallel and stable π-π stacks on the active catalytic site [3]. Other than that, coumarins in both natural and synthesized exhibit wide range of bioactivities including antioxidant which is essential to address one of the contributing factors of Alzheimer's disease that is oxidative stress [21][22][23]. The second moiety that can interact with the other active site is the benzyl amino group. Benzyl amino group was present in many potent AChE inhibitors and its binding capability at the cent re of AChE is demonstrated by the X-ray crystallographic studies of the AChE/donepezil and AChE/galantamine complexes [18]. Besides, previous systematic reviews found that the most potent AChE inhibitor comparable with rivastigmine possess benzyl amino group as one of its moieties [20].
In this study, selenazole ring and urea moiety were chosen as the spacer. Selenazole ring is an aromatic ring and may interact with the aromatic residues located at the AChE gorge through - stacking [18]. Besides, selenazole ring w as selected because of the presence of selenium, a known antioxidant that can reduce oxidative stress, which is a known factor of Alzheimer's disease [3,24,25]. While urea moiety was selected for its ability to form H-bonds, forming complexes and its devouring extensive biological activity [3]. Therefore, the combination of selenazole ring and urea as the spacer is predicted to increase the interaction with the residues lining, the wall of AChE gorge. Thus, the antioxidant activity of the compounds can be improved as a whole.
The new hybrid molecule with a modified coumarin structure obtained by replacing the thiazole ring with selenazole ring (blue coloured) was demonstrated (Scheme 1). This compound bears a resemblance of the potent AChE inhibitor, AP2238 [18]. A broad approach in determining the potential of 1,3-selenazole hybrid coumaryl compound as an AChE and BChE inhibitor via in vitro testing and molecular docking are also highlighted in this work. To the best of our knowledge, there is a limited study on the synthesis and biological activities of coumaryl 1,3 -selenazole structures. This is a continuation of previous work on the synthesis of coumaryl 1,3 -selenazole derivatives and its biological activities and the crystal structure of a coumaryl 1,3-selenazole [23,26].

Materials and Methods
All the new compounds were characterized by 1 H NMR, 13 C NMR, IR and electron spray ionization-mass spectrometry. In the infrared spectra of the synthesized compounds, it was possible to observe the absorption 2400-3400 cm -1 for OH group of carboxylic acid for maleamic acid derivatives. Absorption between 3231 and 3488 cm -1 observed related to NH stretch of urea derivatives while between 3180 and 3300 cm -1 recorded for maleamic acid derivatives. The absorption between 1524 and 1566 cm -1 corresponded to the C=N stretch for selenazole for both derivatives. The absorption for the 6 of 22 C=O group was observed between 1725 and 1737 cm -1 from coumarin carbonyl moiety stretch and between 1692 and 1702 for urea carbonyl moiety stretching. The maleamic acid derivatives ha d three absorptions for C=O group observed between 1641 and 1739 cm -1 representing the carbonyl group for carboxylic acid, coumarin and amide groups.
From the 1 H NMR spectrum, two signals due to the hydrogen attached to the amide nitrogen were observed at 8.26 and 10.67 ppm for urea derivatives for compound 5a. The only proton at selenazole ring was detected at 8.60 ppm. This result is supported by the literature data [ [30]]. The signals for the vinyl proton of maleamic acid derivatives, 6a was observed at 6.51 and 6.56 ppm. From the 13 C NMR, the signal at 157.2 to 159.4 ppm was assigned to the coumarin carbonyl for both derivatives, 5a-5d and 6a-6f.
On the other hand, the signal at 151.4 to 151.8 ppm was assigned to the carbonyl group for urea. The carbon at selenazole ring which attached to the urea or amide moiety is the most downfield signal observed at 160.3 ppm for 5a and 167.7 ppm for 6a. The C=O signal for amide and carboxylic acid for 6a-6f compounds were observed at 161.2 and 167.8 ppm, respectively.

Cupric Ion Reducing Antioxidant Capacity (CUPRAC) assay
CUPRAC is a better antioxidant assay compared to other electron-transfer based assay due to its realistic pH close to physiological pH, favourable redox potential, accessibility and stability of reagents and applicability to lipophilic antioxidants as well as hydrophilic ones. The CUPRAC assay of the synthesized compounds was determined as described previously using Trolox as the reference compound [31]. All synthesized compounds showed higher A0.50 values compared to Trolox (A0.50= 132.74 μM). Compounds 5a-5c exhibited A0.50 values >1000 μM, which showed that these compounds have weak antioxidant activity.

AChE and BChE inhibitory activities
AChE and BChE inhibitory activities of the compounds determined based on the work of Jamila et al. [32] by employing physostigmine as the reference compound. IC50 values of AChE and BChE inhibitions are summarized in Table 2. Overall, the results indicated that most of the compounds exhibited moderate inhibition against both enzymes. The IC50 values of the compounds were recorded between 56.01 and 221.35 μM for AChE inhibition assay. Meanwhile, the range of IC50 value for the compounds in the BChE inhibitory assay was found between 121.34 and 311.37 μM. The selectivity index showed that all the compounds tested were not selective towards AChE or BChE. However, statistical differences in AChE and BChE activities were detected for all test compounds except for compound 5c.
The IC50 value for AChE increased from 124.12 μM to 183.74 μM when Cl (compound 5a) is substituted with Br (compound 5b) at C-16. The increment showed that 5b has lower inhibition potency. This observation differs from Kurt et al. [3] which reported that a larger substituent on a phenyl ring exhibits better inhibition activity. The addition of two Cl atom at C-15 and C-16 in compound 6f caused a decrease in IC50 value for AChE and an increase in IC50 value for BChE as compared to compound 6a which does not contain a Cl atom in its structure. Compound 6d and 6f have the same molecular weight because both compounds contain two Cl atoms but located at different carbon atoms. Modification of Cl atom position from the coumarin ring (6d) to C-15 and C-16 (6f) caused a drastic decrease in IC50 BChE. This shows that the Cl atom is more suitable as a substituent at C-15 than C-16 for BChE inhibition. The Kurt's sulfur analogue [3], compound 5 series with the same skeletal structure gives IC50 values between 31.46 and > 200 μM for AChE, while the IC50 values for BChE are between 4.93 and 194.55 μM. This showed that the substitution of the 1,3-selenazole moiety did not exhibit any differences in inhibition activity of AChE and BChE as compared to 1,3 -thiazole moiety reported by Kurt et al. [3]. This is probably because the synthesized compounds are bigger than previous work as it contains selenium atom which has bigger atomic size than sulfur making it challenging to bind in the binding pocket of AChE [3,20]. Table 2.

Docking studies and binding mode analysis
The most active compounds, 5c and 6b, were selected to further examined their binding modes in binding pockets of both cholinesterase while donepezil was used as a reference drug for the docki ng studies. Table 3 summarizes the XP GScore values and binding site interactions between compound 5c, 6b and donepezil with AChE. In human acetylcholinesterase (hAChE), phenyl ring of donepezil formed - stacking with Trp 286 in the peripheral site of the active site and Trp 86 in the choline-binding pocket of active site similar to previous studies [34]. Besides, the carbonyl group of donepezil formed H-bonding with Phe 295. Meanwhile, compounds 5c and 6b showed good binding score against hAChE, which were -10.1 and -8.9 kcal/mol respectively.
However, the binding scores of both compounds were higher than donepezil (-16.8 kcal/mol). Interestingly, compound 5c was found to simultaneously interact with the peripheral site and catalytic site (acyl binding pocket) of hAChE. The carbonyl group of lactone group capable of forming Hbonding with essential residue, Phe 295 at the acyl binding pocket which was similar to donepezil. The binding interactions of compound 5c were further strengthened with the formation of parallel π-π stacking between the phenyl ring of coumarin and Tyr 341 at the peripheral site ( Figure 2). In contrast, compound 6b only showed interaction at the peripheral site. The lactone group formed parallel π-π stacking with Tyr 341 while C=O of carboxylic acid formed H-bonding with Tyr 124 (Figure 3).
Moreover, compound 6b fitted well in the gorge of hAChE as compared to compound 5c, which did not fit as well. It was probably due to the presence of a large substituent group (Br) at the phenyl ring. Both compounds also exhibited π-π stacking interactions with residues of AChE. However, this interaction did not involve Trp 286, a residue at the peripheral site, and Trp 86, a residue at catalytic cleft which are crucial in achieving strong inhibitory effects by dual binding site inhibition.   On the other hand, the binding scores for compounds 5c and 6b against hBChE were -6.6 dan -6.3 kcal/mol, respectively (Table 4). These values were slightly higher than the binding score of donepezil against BChE, which was -7.6 kcal/mol. Both compounds showed π-π stacking with essential residues of hBChE [34]. Compound 5c exhibited T-shaped π-π stacking between the coumarin phenyl ring and both Trp 231 and Phe 329 residues of the acyl binding pocket (Figure 4) while compound 6b showed Tshaped π-π stacking between the Tyr 332 residue and selenazole ring at the peripheral site ( Figure 5).
Donepezil showed simultaneous π-π stacking at acyl binding pocket (Trp 231) and choline-binding pocket (Trp 82). The interaction with essential residues of hBChE which are Trp 231 and Trp 329 at acyl binding pocket, Tyr 332 at the peripheral site and Trp 82 at choline-binding pocket [35], suggested that compounds 5c and 6b can be considered as a potential BChE inhibitor .

ADMET profile of compounds 6a-6e
In general, all compounds exhibit a good level of human intestinal absorption and aqueous solubility. Table 5 summarizes the ADMET profiles for compounds 6a-6f. In addition, compounds 6a-6b show low blood-brain barrier (BBB) penetration while the rest are undefined. In term of protein plasma binding properties capability, all the synthesized compounds are predicted to be binders.
Further, all compounds are non-inhibitors to the cytochrome P450 2D6 enzyme. Finally, the in-silico prediction revealed that compounds 6a-6e are generally hepatotoxic. The intriguing results open further investigation for future works on improving the toxicity profile of the compounds. e Level of aqueous solubility prediction; 0 (extremely low), 1 (very low), 2 (low), 3 (good), 4 (optimal), 5 (too soluble), 6 ( warning: molec ules with one or more unknown Alog P calculation).

Materials and Methods
Starting materials and chemical reagents were purchased from Sigma -Aldrich and Merck, and they were used without purification. Tetrahydrofuran (THF) was dried and distilled before use. The chemical reactions were routinely checked on Merck TLC plate silica gel 60 F254 in every reaction step. Purification procedures were conducted using column chromatography on Merck silica gel 60 (mesh 230-400). Melting points were taken on an Electrothermal 9100. IR spectra were measured on a Perkin Elmer Spectrum 400 FT-IR/FT-NIR with spotlight 400 Imaging system spectrometer. 1 H and 13 C NMR spectra were measured on a Bruker Avance III HD at 400 and 100 MHz, respectively. Mass spectra were obtained using Agilent 7890A. Electrospray ionization mass spectrometry was conducted using Dionex Ultimate 3000 and Bruker Daltonic/MicroTOF Q. Spectrophotometric analysis was performed using an EPOCH (Biotek, USA) for antioxidant assay and SPECTROstar Nano BMG LabTech for inhibitory activities of AChE and BChE.

General procedure for the synthesis of compound 4a-4e
The appropriate 3-(2-bromoacetyl)-2H-chromen-2-one (1 mmol) and selenourea (1 mmol) were dissolved in methanol: water (2 mL: 2 mL) containing 0.02 g of NaF. The mixture was magnetically stirred at room temperature for 15 to 30 minutes. After completion of the reaction, 5 mL of water was added, and the crude product was filtered and washed with water. Then, the crude product was extracted with ethyl acetate to yield pure substituted 1,3-selenazoles [29,36].

General procedure for the synthesis of compound 5a-5d
Isocyanate derivatives (1 mmol) were added to a solution of 4a-4e (1 mmol) and triethylamine (1 mL) in dry THF. The mixture was refluxed under N2 atmosphere for 12 hours with stirring, then cooled and evaporated to dryness. The crude product was washed wit h chloroform and dried under vacuum [3]. The products were recrystallized from THF to obtained 40-60 % yields.

General procedure for the synthesis of compound 6a-6f
Maleic anhydride (2 mmol) was dissolved in THF. Then, compound 4a-4d (2 mmol) w as added, and the mixture w as magnetically stirred for 3 hours. The precipitate was filtered to give a pure product with 86-100 % yield.

Antioxidant activity assays
In CUPRAC assay, the absorbance values were used to calculate for A0.50, but in DPPH and ABTS assay, inhibition (%) values were used to calculate for IC50.
4.6.1. DPPH a radical scavenging assay The DPPH radical scavenging assay of the derivatives w as measured according to the method reported previously with some modification [37]. Briefly, 1.5 mL aliquot of each sample at 16.125, 31.25, 62.50, 125 and 250 µg/mL was added to 1 mL of 0.1 mM DPPH in methanol. The mixture was agitated vigorously for 1 min and allowed to stand in the dark for 90 min at room temperature. The absorbance value was recorded at 517 nm. Gallic and ascorbic acids were used as a reference control. All measurements were carried out in triplicate. The radical scavenging assay of samples was expressed as percentage inhibition of DPPH using the following equation (1): Where Ac is the absorbance value of the control (DPPH solution without samples); and A0 is the absorbance value of the compound (DPPH solution with samples). 4.6.2. ABTS cation radical scavenging assay ABTS scavenging activities of the synthesized compounds were determined according to the literature method [38]. The solution of ABTS radical was generated by dissolving 19.2 mg of 2,2ʹ -azino-bis(3ethylbenzothiazoline-6-sulphonic acid)(7 mM ABTS) and 3.3 mg K2S2O8 in distilled water (5 mL). The solution kept in the dark for 24 h at room temperature and their absorbance fixed to 0.70 (± 0.02) at 734 nm by dilution. The stock solutions of the samples were prepared in methan ol or 2-propanol at a concentration of 1000 μg/ mL. Then the samples were diluted to 500, 250, 125 and 62.5 μg/ mL. The absorbance was measured in room temperature at 734 nm after 10 minutes from ABTS addition. All measurements were carried out in triplicate. The results were calculated as IC50.

CUPRAC assay
Cupric reducing antioxidant capacities of the synthesized compounds were determined according to the literature method [31]. The sample solutions and Trolox (standards) were prepared in methanol at a concentration of 1000 μg/ mL. Then, sample solutions were diluted to 500, 250, 125, 62.5, 31.25 μg/ mL. Trolox solutions were diluted in the range of 1.0 -62.5 μg/ mL. The addition of the solution is as follows: 1 mL Cu(II) + 1 mL neocuproin + 1 mL buffer solution + 1 mL sample/standard + 0.1 mL H2O = 4.1 mL. The absorbance was measured in room temperature at 450 nm after an hour. The res ults were calculated as A0.50. Methanol was used as a solvent to controls.

AChE and BChE inhibition assay
Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 20 August 2020 doi:10.20944/preprints202008.0438.v1 AChE and BChE inhibitory activity of the compounds w as evaluated using Ellman's spectrometric assay following the conditions as described by Jamila et al. [32]. Electric eel AChE (Sigma Code: C2888) and equine serum BChE (Sigma Code: C7512) were used as the sources of cholinesterases, while acetylthiocholine iodide (Sigma Code: A5751) and S-butyrylthiocholine iodide (Sigma Code: 20820) were used as the sources of substrates. The enzymes were prepared at a concentration of 0.09 U/mL in 0.1 M sodium phosphate buffer (pH 7.4). 14 mM ATCI/ BTCI solution was prepared in deionized water while 10 mM DTNB (Sigma Code: D218200) solution was prepared in phosphate buffer solution. Physostigmine (Sigma Code: E8375) was prepared in methanol at the initial concentration of 1 mg/mL and were diluted between 0 to 100 μg/mL. Sample solutions were prepared by dissolving the sample in methanol or DMSO at the initial concentration of 1 mg/mL, and it was diluted in a phosphate buffer solution to the concentration of 2.5 -100 μg/mL. Next, 140 μL of phosphate buffer solution, 20 μL of test compounds or solvent and 20 μL of the enzyme w as firstly added to 96-well microplate. The mixture was incubated at room temperature for 15 minut es. Then, 10 μL DTNB and 10 μL ATCI/BTCI was added to initiate the enzymatic reaction. The absorbance was measured at 412 nm using a microplate reader (SPECTROstar Nano, BMG LabTech, Offenburg, Germany). All reactions were carried out in triplicate. The IC50 of compounds were calculated in μM.

Ligands preparation
The 2D structures of compounds 5c and 6c were drawn using MarvinSketch (ChemAxon Ltd., Budapest, Hungary) and converted to 3D structures in Maestro (Schrӧdinger, LLC, New York, USA). Before docking, these structures were geometrically optimized and minimized us ing OPLS3 in LigPrep (Schrӧdinger, LLC, New York, USA).

Receptors preparation
The crystal structures of human AChE (hAChE; PDB Code: 4EY7) and human BchE (hBChE; PDB Code: 5DYW) were obtained from the Protein Data Bank (PDB). The cholinesterase str uctures were prepared and optimized using the Protein Preparation Wizard (Schrӧdinger, LLC, New York, USA). All water molecules and cofactors were removed while mutation of incomplete amino acid residues and addition of missing residues was computed through Maestro and followed by minimizing the energy using Optimized Potentials for Liquid Simulations (OPLS3). Then, a receptor grid was generated around the active site of AChE and BChE by choosing centroid of co-crystallized ligands of each enzyme which is donepezil for 4EY7 and N-{[(3S)-1-benzylpiperidin-3-yl]methyl}-N-(2-methoxyethyl)naphthalene-2sulfonamide for 5DYW. Grid box size was set to 20 Å radius, using receptor grid generation implemented in Glide (Schrӧdinger, LLC, New York, USA).

Molecular docking
Molecular docking was carried out using Glide (Schrӧdinger, LLC, New York, USA). All docking calculations were performed using Extra Precision (XP) mode, and the best-docked complex of each compound was determined based on the Glide docking score. The interactions of the docked structures were further analyzed and visualized using PyMOL (Schrӧdinger, LLC, New York, USA).

Data analysis
Data were subjected to one-way analysis of variance (ANOVA) using SPSS software package for windows (IBM SPSS Statistics 25). The IC50 (AChE) values were obtained from the plotted data using Microsoft Excel for Windows (Office 365). This study used α = 0.05. If the P-value ≤ α, there is enough evidence to conclude that the mean of IC50 is different for the two compounds.

Conclusions
In summary, a series of novel coumaryl 1,3-selenazole derivatives (5a-5d and 6a-6f) were synthesized, and their antioxidant activities and effects on AChE and BChE were evaluated in this study. Compounds tested demonstrated weak antioxidant activities and weak inhibition against AChE and BChE. The IC50 value for DPPH radical scavenging assay, ABTS cation radical scavenging assay and CUPRAC were >300 μM, >1000 μM and >200 μM respectively , which were high compared to the standards. The IC50 values for AChE were reported between 56.01 and 221.35 μM, and the BChE value was observed, ranging from 121.34 to 311.37 μM. These IC50 values were higher compared to positive controls (physostigmine) which exhibit IC50 value of 10.12 μM for AChE and 89.06 μM for BChE. This is probably because the synthesized compounds are larger than previous work as it contains selenium which has bigger atomic size than sulfur making it challenging to bind in the binding pocket of AChE. However, the molecular docking studies of the most active inhibitor in each series, 5c and 6b compounds revealed that both displayed essential interactions with hAChE and hBChE. The overall findings suggested that coumaryl and phenyl moiety are crucial in forming π -π stacking with residues of hAChE and hBChE. Besides, the target moiety, selenazole ring, showed interaction only with residues of hBChE.