Replacement of Chalcone-Ethers with Chalcone-Thioethers as Potent and Highly Selective Monoamine Oxidase-B Inhibitors and Their Protein-Ligand Interactions

To develop new potent and highly selective MAO-B inhibitors from chalcone-thioethers, eleven chalcones-thioethers were synthesized and their monoamine oxidase (MAO) inhibition, kinetics, reversibility, and cytotoxicity of lead compounds were analyzed. Molecular dynamics were carried out to investigate the interactions. Compound TM8 showed potent inhibitory activity against MAO-B, with an IC50 value of 0.010 µM, followed by TM1, TM2, TM7, and TM10 (IC50 = 0.017, 0.021, 0.023, and 0.026 µM, respectively). Interestingly, TM8 had an extremely high selectivity index (SI; 4860) for MAO-B. Reversibility and kinetic experiments showed that TM8 and TM1 were reversible and competitive inhibitors of MAO-B with Ki values of 0.0031 ± 0.0013 and 0.011± 0.001 µM, respectively. Both TM1 and TM8 were non-toxic to Vero cells with IC50 values of 241.8 and 116.3 µg/mL (i.e., 947.7 and 402.4 µM), respectively, and at these IC50 values, both significantly reduced reactive oxygen species (ROS) levels. TM1 and TM8 showed high blood-brain barrier permeabilities in the parallel artificial membrane permeability assay. Molecular dynamics studies were conducted to investigate interactions between TM1 and TM8 and the active site of MAO-B. Conclusively, TM8 and TM1 are potent and highly selective MAO-B inhibitors with little toxicity and good ROS scavenging abilities and it is suggested that both are attractive prospective candidates for the treatment of neurological disorders.


Introduction
Parkinson's disease (PD) is considered the second most common progressive neurodegenerative disorder in the elderly and is characterized by motor system disruption [1]. Over ten million people worldwide suffer from PD, which places an onerous responsibility on the medical community to develop an effective therapy [2]. This motor system disruption is caused by the extensive degeneration of the dopaminergic neurons, which leads dopamine (DA) depletion [3]. Current treatment strategies center on enhancing dopamine-like activities using monoamine oxidase (MAO)-B inhibitors, dopamine agonists, and catechol-O-methyltransferase (COMT) inhibitors, whereas developmental efforts appear to be focused on the identification of non-dopaminergic candidates like glutamate antagonists and anticholinergic drugs [4].
MAO is found in two isoforms, that is, MAO-A and MAO-B, and contains flavin adenine dinucleotide (FAD) as a cofactor. These isoforms share 70% amino acid sequence similarity and differ functionally in terms of their substrate specificities and inhibitor affinities [5]. Selective MAO-A and MAO-B inhibitors are considered important for the treatment of various neuropsychiatric and neurodegenerative disorders, respectively [6]. For example, the level of endogenous dopamine enhances the blockade of dopamine metabolism by selective MAO-B inhibition [7]. Numerous studies have documented that the formation of the 1-methyl-4-phenylpyridinium ion (MPP + ; a bioactive metabolite of 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)) is catalyzed by MAO-B, and that MPP + induces the neurotoxic effects associated with the pathophysiology of PD [8]. MPP + interacts with mitochondrial complex 1 to induce neural toxicity in PD [9]. In view of the important roles played by MAO-B in neurodegenerative disorders, novel highly selective MAO-B inhibitors are of considerable interest [10].
A recent strategy used to produce highly selective MAO-B inhibitors included the installation of two aryl or heteroaryl rings on either end of the two to four carbons or nitrogen-containing spacer units, which include α-β-unsaturated ketones, pyrazolines, enamides, and hydrazones [11][12][13][14][15]. The presence of a π-conjugated system with rotatable bonds in the linker can provide different orientations and facilitate interactions between designed scaffolds and the active site of MAO-B [16]. Chalcones are considered versatile scaffolds for selective MAO-B inhibitors. Recently, Guglielmi et al. predicted that chalcone derivatives are likely to exhibit selective human MAO-B (hMAO-B) inhibition based on results obtained using the free online tool SwissTargetPrediction (www.swisstargetprediction.ch (accessed on 10 October 2021)), which supports our previous experimental results [17,18].
Chalcones are α-β unsaturated ketones with three rotatable bonds and are endowed with a number of central nervous system (CNS) related activities, which include adenosine receptor antagonist and neuroprotective activities, ability to cross the blood-brain barrier (BBB), and inhibitions of MAO-B, acetylcholine esterase (AChE), transglutaminase, and β-amyloid plaques formation [19][20][21][22]. Numerous studies have demonstrated that substituent selection and the orientations of substituents on the chalcone A and/or B rings provide a promising means of producing reversible and selective inhibitors of hMAO-A and -B. Notably, the presence of lipophilic electron-donating groups (i.e., ethyl, methoxy, methyl, and dimethylamino, or halogens) at the same position on the B ring results in hMAO-B inhibition in the nanomolar range [23][24][25][26], whereas the introduction of methoxy at para-position of the A ring is associated with hMAO-B inhibitory activity [27][28][29][30]. The design strategy used in the present study involved the isosteric replacement of the A ring methoxy with a thiomethyl group and the introduction of different substituents on the B ring (Scheme 1). Scheme 1. Synthetic route to chalcone-thiomethyl ethers TM1-11. Scheme 1. Synthetic route to chalcone-thiomethyl ethers TM1-11.
To the best of our knowledge, the influence of isosteric replacement of the A-ring methoxy with a thiomethoxy group has not been previously reported to enhance hMAO-B inhibition ( Figure 1).
To the best of our knowledge, the influence of isosteric replacement of the A-ring methoxy with a thiomethoxy group has not been previously reported to enhance hMAO-B inhibition (Figure 1). We report a series of chalcone-thioether derivatives, which were prepared and evaluated for their MAO-A and MAO-B inhibitory activities, cytotoxicities, and effect on reactive oxygen species (ROS). Lead molecules were subjected to detailed molecular dynamics (MD) study to access their molecular mechanisms.

MAO Inhibition
All eleven chalcone derivatives potently inhibited MAO-B at 10 µM; residual activities were <15% (Table 1). Compound TM8 showed the highest inhibitory activity against MAO-B with an IC 50 value of 0.010 µM, and TM1, TM2, TM7, and TM10 also exhibited potent inhibitory effects (IC 50 = 0.017, 0.021, 0.023 and 0.026 µM, respectively). The other six compounds had IC 50 values < 0.15 µM. Interestingly, TM1, TM2, TM4, TM7, TM8, and TM10 were 3.7, 3.0, 1.1, 2.7, 6.3 and 2.4 times, respectively, more potent than the reference drug lazabemide. Similarly, TM1, TM2, and TM8 were 1.6, 1.3 and 2.8 times, respectively, more potent than pargyline. Compounds TM7 and TM10 were almost as potent as pargyline. Regarding the effects of substituents, the inhibitory potencies of these compounds followed the order -Cl > -H > -OH > -NO 2 > -F. To the best of our knowledge, the influence of isosteric replacement of the A-ring methoxy with a thiomethoxy group has not been previously reported to enhance hMAO-B inhibition (Figure 1). We report a series of chalcone-thioether derivatives, which were prepared and evaluated for their MAO-A and MAO-B inhibitory activities, cytotoxicities, and effect on reactive oxygen species (ROS). Lead molecules were subjected to detailed molecular dynamics (MD) study to access their molecular mechanisms.

MAO Inhibition
All eleven chalcone derivatives potently inhibited MAO-B at 10 µ M; residual activities were <15% (Table 1). Compound TM8 showed the highest inhibitory activity against MAO-B with an IC50 value of 0.010 µ M, and TM1, TM2, TM7, and TM10 also exhibited potent inhibitory effects (IC50 = 0.017, 0.021, 0.023 and 0.026 µ M, respectively). The other six compounds had IC50 values < 0.15 µ M. Interestingly, TM1, TM2, TM4, TM7, TM8, and TM10 were 3.7, 3.0, 1.1, 2.7, 6.3 and 2.4 times, respectively, more potent than the reference drug lazabemide. Similarly, TM1, TM2, and TM8 were 1.6, 1.3 and 2.8 times, respectively, more potent than pargyline. Compounds TM7 and TM10 were almost as potent as pargyline. Regarding the effects of substituents, the inhibitory potencies of these compounds followed the order -Cl > -H > -OH > -NO2> -F. To the best of our knowledge, the influence of isosteric replacement of the A-ring methoxy with a thiomethoxy group has not been previously reported to enhance hMAO-B inhibition (Figure 1). We report a series of chalcone-thioether derivatives, which were prepared and evaluated for their MAO-A and MAO-B inhibitory activities, cytotoxicities, and effect on reactive oxygen species (ROS). Lead molecules were subjected to detailed molecular dynamics (MD) study to access their molecular mechanisms.

MAO Inhibition
All eleven chalcone derivatives potently inhibited MAO-B at 10 µ M; residual activities were <15% (Table 1). Compound TM8 showed the highest inhibitory activity against MAO-B with an IC50 value of 0.010 µ M, and TM1, TM2, TM7, and TM10 also exhibited potent inhibitory effects (IC50 = 0.017, 0.021, 0.023 and 0.026 µ M, respectively). The other six compounds had IC50 values < 0.15 µ M. Interestingly, TM1, TM2, TM4, TM7, TM8, and TM10 were 3.7, 3.0, 1.1, 2.7, 6.3 and 2.4 times, respectively, more potent than the reference drug lazabemide. Similarly, TM1, TM2, and TM8 were 1.6, 1.3 and 2.8 times, respectively, more potent than pargyline. Compounds TM7 and TM10 were almost as potent as pargyline. Regarding the effects of substituents, the inhibitory potencies of these compounds followed the order -Cl > -H > -OH > -NO2> -F. To the best of our knowledge, the influence of isosteric replacement of the A-ring methoxy with a thiomethoxy group has not been previously reported to enhance hMAO-B inhibition (Figure 1). We report a series of chalcone-thioether derivatives, which were prepared and evaluated for their MAO-A and MAO-B inhibitory activities, cytotoxicities, and effect on reactive oxygen species (ROS). Lead molecules were subjected to detailed molecular dynamics (MD) study to access their molecular mechanisms.

Kinetics
Kinetic studies on MAO-B inhibitions were performed on TM1 and TM8. Lineweaver-Burk plots and secondary plots showed that the lines intersected y-axes, thus indicating, TM1 and TM8 competitively inhibited MAO-B (Figure 2A,C). K i values were 0.011 ± 0.001 and 0.0031 ± 0.0013 µM, respectively ( Figure 2B,D). These results suggest that compounds TM1 and TM8 compete with the substrate at the active site of MAO-B.

Kinetics
Kinetic studies on MAO-B inhibitions were performed on TM1 and TM8. L eweaver-Burk plots and secondary plots showed that the lines intersected y-axes, th indicating, TM1 and TM8 competitively inhibited MAO-B (Figure 2A,C). Ki values w 0.011 ± 0.001 and 0.0031 ± 0.0013 µM, respectively ( Figures 2B,D). These results sugg that compounds TM1 and TM8 compete with the substrate at the active site of MAO-B

Reversibility
MAO-B inhibition reversibility studies were conducted on compounds TM1 a TM8 using a dialysis-based method. MAO-B inhibition by TM1 was recovered by dialy from 24.9% (AU) to 68.1% (AD), and inhibition by TM8 was recovered from 33.1% to 69.3 These recoveries were similar to that observed for the reference reversible lazabemi (from 16.6% to 74.9%). Inhibition of MAO-B by pargyline (the reference irreversible inh itor) was not recovered (from 24.7% to 31.4%) (Figure 3). These results showed that co pounds TM1 and TM8 were reversible MAO-B inhibitors.

Reversibility
MAO-B inhibition reversibility studies were conducted on compounds TM1 and TM8 using a dialysis-based method. MAO-B inhibition by TM1 was recovered by dialysis from 24.9% (A U ) to 68.1% (A D ), and inhibition by TM8 was recovered from 33.1% to 69.3%. These recoveries were similar to that observed for the reference reversible lazabemide (from 16.6% to 74.9%). Inhibition of MAO-B by pargyline (the reference irreversible inhibitor) was not recovered (from 24.7% to 31.4%) (Figure 3). These results showed that compounds TM1 and TM8 were reversible MAO-B inhibitors.

Cytotoxicity Evaluation
African Green Monkey kidney cells (VERO cells) were used to evaluate the cytotoxic effects of TM1 and TM8. Cells were incubated with different doses (1 µg/mL to 500 µg/mL) of the two compounds for 48 h, and relative cell viabilities were determined by MTT assay. Compound TM1 and TM8 reduced viabilities concentration-dependently (Figures 4a and 5a); TM1 and TM8 treatments resulted in high viability (>90%) at concentrations <250 and <100 µg/mL (i.e., 983.1 and 346.3 µM), respectively, and the IC50 values were 241.6 and 116.2 µg/mL (i.e., 947.7 and 402.4 µM), respectively (Figures 4b and 5b). Phase contrast microscopy showed that treatments above IC50 values resulted in membrane damage and reduced cell numbers. These results suggest that TM1 and TM8 are safe at low concentrations (Figures 4c and 5c), and that at concentrations > their IC50 values induced oxidative stress.

Cytotoxicity Evaluation
African Green Monkey kidney cells (VERO cells) were used to evaluate the cytotoxic effects of TM1 and TM8. Cells were incubated with different doses (1 µg/mL to 500 µg/mL) of the two compounds for 48 h, and relative cell viabilities were determined by MTT assay. Compound TM1 and TM8 reduced viabilities concentration-dependently (Figures 4a and 5a); TM1 and TM8 treatments resulted in high viability (>90%) at concentrations <250 and <100 µg/mL (i.e., 983.1 and 346.3 µM), respectively, and the IC50 values were 241.6 and 116.2 µg/mL (i.e., 947.7 and 402.4 µM), respectively (Figures 4b and 5b). Phase contrast microscopy showed that treatments above IC50 values resulted in membrane damage and reduced cell numbers. These results suggest that TM1 and TM8 are safe at low concentrations (Figures 4c and 5c), and that at concentrations > their IC50 values induced oxidative stress.

ROS Assay
ROS increase vulnerability to neuronal damage by causing oxidative damage in the brain during neurodegenerative diseases. When Vero cells were exposed to exogenous H 2 O 2 , intracellular ROS levels were elevated, but treatments with TM1 or TM8 at their IC 50 values resulted in appreciably lower ROS levels (Figure 6a,b).

ROS Assay
ROS increase vulnerability to neuronal damage by causing oxidative damage in the brain during neurodegenerative diseases. When Vero cells were exposed to exogenous H2O2, intracellular ROS levels were elevated, but treatments with TM1 or TM8 at their IC50 values resulted in appreciably lower ROS levels (Figure 6a,b). were pre-incubated with 100 µg/mL of 30% H2O2 and ROS production was evaluated immediately as described in Section 4.

Blood-Brain Barrier (BBB) Permeation Study by Parallel Artificial Membrane Permeability Assay (PAMPA)
A highly effective permeability and high CNS bioavailability of thiomethane-containing chalcones were demonstrated with Pe ranges between 16.26 and 13.28 × 10 −6 cm/s (Table 2) in the parallel artificial membrane permeability assay (PAMPA).

ROS Assay
ROS increase vulnerability to neuronal damage by causing oxidative damage in the brain during neurodegenerative diseases. When Vero cells were exposed to exogenous H2O2, intracellular ROS levels were elevated, but treatments with TM1 or TM8 at their IC50 values resulted in appreciably lower ROS levels (Figure 6a,b). Figure 6. Effects of TM1 (a) and TM8 (b) on H2O2-induced ROS levels. Vero cells (1 × 10 6 cells/mL) were pre-incubated with 100 µg/mL of 30% H2O2 and ROS production was evaluated immediately as described in Section 4.

Blood-Brain Barrier (BBB) Permeation Study by Parallel Artificial Membrane Permeability Assay (PAMPA)
A highly effective permeability and high CNS bioavailability of thiomethane-containing chalcones were demonstrated with Pe ranges between 16.26 and 13.28 × 10 −6 cm/s (Table 2) in the parallel artificial membrane permeability assay (PAMPA).

Blood-Brain Barrier (BBB) Permeation Study by Parallel Artificial Membrane Permeability Assay (PAMPA)
A highly effective permeability and high CNS bioavailability of thiomethane-containing chalcones were demonstrated with Pe ranges between 16.26 and 13.28 × 10 −6 cm/s (Table 2) in the parallel artificial membrane permeability assay (PAMPA).

Inhibitor-Induced Binding Pocket Dynamics
Binding pocket residues of MAO-B exhibited average root mean square deviations (RMSDs), of 1.77, 1.81, 1.29, 1.64 and 1.81 Å for TM1, TM2, TM7, TM8 and TM10, respectively (Figure 7), suggesting relatively stable conformations since average values were below 2 Å. An average RMSD of 2.59 Å was estimated for unbound MAO-B, which suggested a relatively unstable conformation. According to Luque and Freire in 2000 [19], a stable binding pocket conformation provides optimum binding affinity for small ligands. We assessed the rigidity or compactness of the binding pocket residues by estimating their conformational changes within the binding pocket of MAO-B mediated by TM1, TM2, TM7, TM8 and TM10, which altered the overall structure and function of MAO-B radii of gyration over a 250 ns MD simulation period. Relatively stable conformations were observed, and binding pocket residues in inhibitor-bound MAO-B complexes were more compact/rigid with average radii of gyration (ROGs) of 12.39, 12.56, 12.55, 12.57 and 12.33 Å for TM1, TM2, TM7, TM8 and TM10, respectively. Unbound MAO-B binding pocket residues exhibited a relatively higher average ROG of 12.65 Å, which suggest that bindings of TM1, TM2, TM7, TM8 and TM10 enhanced binding pocket stability. An assessment of the solvent surface exposure of binding pocking residues also provided insights into the conformational dynamics of the binding pocket after inhibitor binding [14,20,21]. As shown in Figure 8

Discussion
Chalcones provide a simple framework with broad-spectrum biological activities and are accessible by Claisen-Schmidt condensation. The aldehyde used becomes the B ring of the chalcone, whereas the acetophenone used becomes the A ring. The α,β-unsaturated ketone moiety of chalcones is considered a Michael acceptor and may be involved in the signaling of many biochemical pathways. The presence of olefinic linkage introduces the possibility of cis-and trans-isomers, but the more thermodynamically stable trans-form is invariably produced. The presence of electron-donating and -withdrawing groups on the phenyl ring of chalcones greatly influences the electron density on the

Discussion
Chalcones provide a simple framework with broad-spectrum biological activities and are accessible by Claisen-Schmidt condensation. The aldehyde used becomes the B ring of the chalcone, whereas the acetophenone used becomes the A ring. The α,β-unsaturated ketone moiety of chalcones is considered a Michael acceptor and may be involved in the signaling of many biochemical pathways. The presence of olefinic linkage introduces the possibility of cisand trans-isomers, but the more thermodynamically stable trans-form is invariably produced. The presence of electron-donating and -withdrawing groups on the phenyl ring of chalcones greatly influences the electron density on the enone moiety and the orientations of the three rotatable bonds of the chalcone framework. These unique features enable this simple framework to interact with enzyme targets, preferably those with hydrophobic pockets.
Our structure-activity relationship (SAR) investigation of the designed chalconethioethers showed reasonable relations with observed inhibitory effects. In the 11 chalcone derivatives, the thioether (-SMe) group was located on the A ring and other alterations were made by placing various substituents at the para-position on the ring B. Compound TM1 (the parent compound) had no substituent at the para-position but had excellent MAO-B inhibitory effect with an IC 50 value of 0.017 ± 0.0026 µM and an SI of 811.8. The K i value of TM1 for MAO-B (0.011 ± 0.001 µM) is much potent than the unsubstituted methoxylated chalcone (C1) previously reported by our group (0.70 ± 0.05 µM) [30]. This crucial observation clearly showed that bio-isosteric replacement with thioether group on the para position of phenyl ring A in the chalcone framework resulted in pronounced MAO-B inhibitory activities. The introduction of chlorine atom on the para position of ring B with thioether group on the A ring resulted highly potent MAO-B inhibition (K i = 0.0031 µM) ( Figure 9). enone moiety and the orientations of the three rotatable bonds of the chalcone framework. These unique features enable this simple framework to interact with enzyme targets, preferably those with hydrophobic pockets. Our structure-activity relationship (SAR) investigation of the designed chalcone-thioethers showed reasonable relations with observed inhibitory effects. In the 11 chalcone derivatives, the thioether (-SMe) group was located on the A ring and other alterations were made by placing various substituents at the para-position on the ring B. Compound TM1 (the parent compound) had no substituent at the para-position but had excellent MAO-B inhibitory effect with an IC50 value of 0.017 ± 0.0026 µM and an SI of 811.8. The Ki value of TM1 for MAO-B (0.011 ± 0.001 µM) is much potent than the unsubstituted methoxylated chalcone (C1) previously reported by our group (0.70 ± 0.05 µM) [30]. This crucial observation clearly showed that bio-isosteric replacement with thioether group on the para position of phenyl ring A in the chalcone framework resulted in pronounced MAO-B inhibitory activities. The introduction of chlorine atom on the para position of ring B with thioether group on the A ring resulted highly potent MAO-B inhibition (Ki = 0.0031µM) ( Figure 9). When -NMe2 (a strong election-releasing group (ERG)) was introduced, resulting TM5 showed a 7-fold fall in inhibitory activity and a 10-fold reduction in selectivity (0.12 ± 0.01 µM and 49.3, respectively). Similar results were observed when -CH2CH3 (a weak ERG) was introduced to produce TM6 (Table 1), which was ascribed to its bulkiness. The results were ascertained by introduction of small but strong and relatively week electronreleasing groups such as -OH, -OMe, and -Me. The -OMe (a strong ERG) increased inhibition and selectivity with respect to TM5 and TM6 (IC50 = 0.088 ± 0.011 µM and 100.2). Of the ERG introduced, those with a weak ERG (e.g., -Me), except -OH, increased inhibition and selectivity (IC50 = 0.058 ± 0.001 µM and 206.9, respectively). The strong ERG -OH, TM2, had an inhibitory effect (IC50 = 0.021 ± 0.003 µM) similar to the parent compound TM1, but considerably less selectivity, which may have been due to hydrogen bonding. Generally, ERGs had marked MAO-B inhibitory effects as compared with the parent compound TM1. The effects of electron-withdrawing groups (EWG) on chalcone-thioethers on MAO-B inhibitory activity were also investigated. The introduction of -NO2 (the strongest EWG) in TM7 resulted in inhibitory activity (IC50 = 0.023 ± 0.001 µM) similar to the parent compound TM1 but improved selectivity (1,304.3), whereas the introduction of -CF3 (a weak EWG) in TM11 caused a considerable drop in inhibitory activity and selectivity (IC50 = 0.081 ± 0.032 µM and 311.1, respectively). Other halogens (also EWGs) such as -F, -Cl, and -Br had effects similar to TM10 (IC50 = 0.026 ± 0.004 µM) and a drop (0.088 ± 0.006 µM) in MAO-B inhibition effect with respect to the parent compound, except -Cl (TM8). In addition, a considerable drop in selectivity was also observed for these groups. The -Cl was a more effective pharmacophore than other halogens and EWGs. TM8 containing-Cl had the highest MAO-B inhibitory activity (IC50 = 0.010 ± 0.005 µM) and selectivity (4860.0) and had 6.3 and 2.8 times more inhibitory activity than the reference drugs lazabemide and pargyline, respectively. In contrast, strong EWGs such as -OH, -OMe, - When -NMe 2 (a strong election-releasing group (ERG)) was introduced, resulting TM5 showed a 7-fold fall in inhibitory activity and a 10-fold reduction in selectivity (0.12 ± 0.01 µM and 49.3, respectively). Similar results were observed when -CH 2 CH 3 (a weak ERG) was introduced to produce TM6 (Table 1), which was ascribed to its bulkiness. The results were ascertained by introduction of small but strong and relatively week electron-releasing groups such as -OH, -OMe, and -Me. The -OMe (a strong ERG) increased inhibition and selectivity with respect to TM5 and TM6 (IC 50 = 0.088 ± 0.011 µM and 100.2). Of the ERG introduced, those with a weak ERG (e.g., -Me), except -OH, increased inhibition and selectivity (IC 50 = 0.058 ± 0.001 µM and 206.9, respectively). The strong ERG -OH, TM2, had an inhibitory effect (IC 50 = 0.021 ± 0.003 µM) similar to the parent compound TM1, but considerably less selectivity, which may have been due to hydrogen bonding. Generally, ERGs had marked MAO-B inhibitory effects as compared with the parent compound TM1. The effects of electron-withdrawing groups (EWG) on chalconethioethers on MAO-B inhibitory activity were also investigated. The introduction of -NO 2 (the strongest EWG) in TM7 resulted in inhibitory activity (IC 50 = 0.023 ± 0.001 µM) similar to the parent compound TM1 but improved selectivity (1,304.3), whereas the introduction of -CF 3 (a weak EWG) in TM11 caused a considerable drop in inhibitory activity and selectivity (IC 50 = 0.081 ± 0.032 µM and 311.1, respectively). Other halogens (also EWGs) such as -F, -Cl, and -Br had effects similar to TM10 (IC 50 = 0.026 ± 0.004 µM) and a drop (0.088 ± 0.006 µM) in MAO-B inhibition effect with respect to the parent compound, except -Cl (TM8). In addition, a considerable drop in selectivity was also observed for these groups. The -Cl was a more effective pharmacophore than other halogens and EWGs. TM8 containing-Cl had the highest MAO-B inhibitory activity (IC 50 = 0.010 ± 0.005 µM) and selectivity (4860.0) and had 6.3 and 2.8 times more inhibitory activity than the reference drugs lazabemide and pargyline, respectively. In contrast, strong EWGs such as -OH, -OMe, -NMe 2 , and -Et exhibited better MAO-A inhibitory effects than the parent compound and other EWGs but had far lower inhibitory effects than the reference drugs (Table 1). SAR in this study can be summarized as shown in Figure 10. NMe2, and -Et exhibited better MAO-A inhibitory effects than the parent compound and other EWGs but had far lower inhibitory effects than the reference drugs (Table 1). SAR in this study can be summarized as shown in Figure 10. Previously, our group developed some methoxylated chalcones as selective MAO-B inhibitors and the lead molecule showed effective inhibitory activity (IC50 = 0.29 ± 0.01 µM) [30]. In current study, the bioisosteric replacement of methoxy group into thiomethyl group showed significantly increased MAO-B inhibition. In another recent study of Shalaby et al., compound (E)-3-(3-bromophenyl)-1-(4-(methylthio)phenyl)prop-2-en-1-one exhibited efficient MAO-B inhibition with IC50 = 0.19 ± 0.01 µM [25]. It is noteworthy that the replacement of methoxy chalcones to thiomethyl resulted in enhancement of overall MAO-B inhibitory profiles with high SI values. Interestingly, the analogues tested resulted in potent inhibitory activities in nanomolar ranges towards MAO-B inhibition.
Toxicity assessment were conducted using Vero cells, which are commonly used to assess viability and lethality at a cellular level. We found that more than 90% of cells treated with TM1 or TM8 at effective concentrations remained viable. However, extensive blebbing and membrane damage were observed when cells were treated at high concentrations. Our cell viability and phase contrast microscopy studies showed that TM1 is a safer proposition than TM8. Furthermore, our investigation of the inhibitory effects of TM1 and TM8 on ROS, which are known to play significant roles in the pathogeneses of neurodegenerative diseases and to cause cell death or oxidative stress (OS) when present in excess, showed TM1 and TM8 at their IC50 levels mitigated H2O2-induced increases in ROS levels.
MD simulation provides a means of exploring the structural changes and functions caused by the binding of small molecules and has been used to investigate associations between protein activities and structural dynamics. By comparing protein conformations in binding pockets before and after ligand binding, we were able to identify inhibitorinduced conformational changes associated with inhibition of MAO-B. To observe these conformational changes, we employed MD simulation metrics, that is, RMSD, ROG, and SASA.
Having established that compound bindings induced conformational changes that might favor binding affinity, we estimated the binding free energy for compound/MAO-B complexes using experimentally determined IC50 values. Using the MM/PBSA approach, Previously, our group developed some methoxylated chalcones as selective MAO-B inhibitors and the lead molecule showed effective inhibitory activity (IC 50 = 0.29 ± 0.01 µM) [30]. In current study, the bioisosteric replacement of methoxy group into thiomethyl group showed significantly increased MAO-B inhibition. In another recent study of Shalaby et al., compound (E)-3-(3-bromophenyl)-1-(4-(methylthio)phenyl)prop-2-en-1-one exhibited efficient MAO-B inhibition with IC 50 = 0.19 ± 0.01 µM [25]. It is noteworthy that the replacement of methoxy chalcones to thiomethyl resulted in enhancement of overall MAO-B inhibitory profiles with high SI values. Interestingly, the analogues tested resulted in potent inhibitory activities in nanomolar ranges towards MAO-B inhibition.
Toxicity assessment were conducted using Vero cells, which are commonly used to assess viability and lethality at a cellular level. We found that more than 90% of cells treated with TM1 or TM8 at effective concentrations remained viable. However, extensive blebbing and membrane damage were observed when cells were treated at high concentrations. Our cell viability and phase contrast microscopy studies showed that TM1 is a safer proposition than TM8. Furthermore, our investigation of the inhibitory effects of TM1 and TM8 on ROS, which are known to play significant roles in the pathogeneses of neurodegenerative diseases and to cause cell death or oxidative stress (OS) when present in excess, showed TM1 and TM8 at their IC 50 levels mitigated H 2 O 2 -induced increases in ROS levels.
MD simulation provides a means of exploring the structural changes and functions caused by the binding of small molecules and has been used to investigate associations between protein activities and structural dynamics. By comparing protein conformations in binding pockets before and after ligand binding, we were able to identify inhibitor-induced conformational changes associated with inhibition of MAO-B. To observe these conformational changes, we employed MD simulation metrics, that is, RMSD, ROG, and SASA.
Having established that compound bindings induced conformational changes that might favor binding affinity, we estimated the binding free energy for compound/MAO-B complexes using experimentally determined IC 50 values. Using the MM/PBSA approach, we calculated the binding energies of the compounds ( Table 1). As shown by the table, the compounds exhibited favorable total binding energies, which was in line with our experimental data. Total binding energies for TM1, TM2, TM7, TM8, and TM10 were −40.83 ± 2.63, −34.40 ± 2.63, −47.02 ± 2.65, −37.02 ± 3.05, and −35.14 ± 2.81 kcal/mol, respectively, which suggested that these chalcone derivatives engage in favorable interactions with the active site residues of MAO-B. Though all the compounds exhibited binding in accord with their free total binding energies, TM7/MAO-B complex had the highest free total binding energy (−47.02 ± 2.65 kcal/mol), which was attributed to greater Van der Waals (VDW) forces and electrostatic (ELE) interactions. Overall, it was noticed that VDW interactions (∆EvdW) were important for ligand binding, whereas the favorable ELE interactions (∆Eele) compensated for unfavorable polar solvation energies.
Intermolecular interactions between compounds and binding site residues influence the stabilities of the complexes formed. The total binding free energies of the compounds TM1 and TM8 were decomposed into individual residue-based contributions using the MM/PBSA method. The VDW and ELE energy contributions were expressed as percentages of total binding free energies to determine which residues and energy constituents had the greatest effects on total energy. As shown in Figure 8, the top three residues that contributed the most to the binding of TM1 against MAO-B were TYR60 (−2.55 kcal/mol), LEU171 (−1.85 kcal/mol), and TYR435 (−1.71 kcal/mol); that to the binding of TM2 were ILE199 (−1.95 kcal/mol), LEU171 (−1.71 kcal/mol), and GLN206 (−1.46 kcal/mol); that to the binding of TM7 were TYR398 (−2.44 kcal/mol), TYR435 (−2.12 kcal/mol), and ILE199 (−2.0 kcal/mol); that to the binding of TM8 were LEU171 (−1.73 kcal/mol), TYR398 (−1.71 kcal/mol) and ILE199 (−1.60 kcal/mol); and that to the binding of TM10 were ILE199 (−2.27 kcal/mol), TYR326 (−1.69 kcal/mol), and ILE316 (−1.61 kcal/mol). Although unfavorable ELE interactions were observed in interaction profiles, as shown by the plots in Figure 8, these would have had little effects on total binding due to the cumulative effects of ELE and VDW energy contributions. Particular focus on the SCH3 group, as highlighted in Figure 8, showed the formation of high-affinity interactions with MAO-B, relative to other moieties thus indicating its cruciality to the binding mechanism of synthesized compounds. Specific interactions of the SCH3 group in each compound included; TM1 (π-sigma and π-alkyl with TYR435, π-alkyl with TYR398), TM2 (π-alkyl with ILE199, LEU164 and PRO104), TM7 (π-sigma with TRP117 and π-alkyl with TRP117 and TRP316), TM8 (π-sigma with TRP119 and π-alkyl with LEU164, LEU171, LEU167) and TM10 (π-alkyl with PRO104). These SCH3-mediated high-affinity interactions could contribute to stability and the overall total binding affinity of each compound.
Study of the residues involved in binding revealed that the comparatively high binding affinity of TM7 could be attributed to relatively high binding energy contributions. Furthermore, these cumulative per-residue energy contributions might explain the MAO-B inhibitory effects of TM1, TM2, TM7, TM8, and TM10. Also, these estimations support predictions based on binding free energies.

Synthesis
Chalcone-thioethers (TM1-11) were synthesized using base-catalyzed Claisen-Schmidt condensation between various para-substituted aromatic aldehydes and 4-thiomethyl acetophenone. Target compounds were produced by stirring ethanol solutions of 0.01 M of 4-thiomethylacetophenone with appropriate aromatic aldehydes containing 40% NaOH solution for 8-10 h [32]. The resulting reaction mixtures were poured into ice-cold water, and the products were collected, washed thoroughly with water, and recrystallized from ethanol. Spectral data were provided in Supplementary Materials.

MAO Enzyme Inhibition
MAO-A activity was estimated continuously for 20 min at 316 nm using 0.06 mM kynuramine as substrate, whereas MAO-B activity was measured for 30 min at 250 nm using 0.3 mM benzylamine as substrate, as described previously [33]. MAO activity assays were performed using recombinant human MAO-A or MAO-B.

Kinetics Study
MAO-A activities were estimated after exposing it to inhibitors at 10.0 µM, whereas MAO-B activities were estimated at an inhibitor concentration of 1.0 µM. IC 50 values were determined by measuring residual enzyme activities. Toloxatone and lazabemide were used as reference reversible inhibitors for MAO-A and MAO-B, respectively, and clorgyline and pargyline as reference irreversible inhibitors. K i values and inhibitor types were determined by kinetic testing, as previously described [34]. Kinetic tests were performed at 5 different substrate concentrations and inhibitor concentrations of 0,~1/2 × IC 50 , IC 50 , and 2 × IC 50 values. Lineweaver-Burk plots (LB) and their secondary plots were used to determine K i values and inhibitor types.

Reversibility Studies
The reversibilities of MAO-B inhibitions by TM1 and TM8 were assessed by dialysis. Test compounds were preincubated with MAO-B at 0.15 µM for 30 min, as described previously [35]. MAO-B was also preincubated with the positive controls for the experiment; lazabemide (a reference reversible MAO-B inhibitor) or pargyline (a reference irreversible MAO-B inhibitor) at 0.20 and 0.30 µM, respectively. Reversibility patterns were determined by comparing the activities of dialyzed (A D ) and undialyzed (A U ) samples.

Cytotoxicity and ROS Assays
The cytotoxicities and ROS quenching abilities of the two lead compounds (TM1 and TM8) were evaluated as previously described [36,37].

BBB Study by PAMPA Method
The CNS bioavailability of the lead candidates was further ascertained by PAMPA method [31].

System Preparation
Initial coordinates of MAO-B (PDB ID: 6RKB) were selected [38]. The structure of MAO-B consists of two identical chains, A and B, which forms a MAO-B homodimer. Each chain contains a cofactor FAD, an inhibitor, water, and non-standard residues. Chain B was removed together with the water and non-standard residues using UCSF Chimera [39], and the 'cleaned' chain A was then saved in pdb format. The 2D structures of the compounds TM1, TM2, TM7, TM8 and TM10 were drawn using Marvin Sketch and optimized with Avogadro to generate 3D structures. Using UCSF Chimera, hydrogen atoms and Gasteiger charges were added and results were saved in mol2 format for molecular docking at the active site of 'cleaned' MAO-B.

Molecular Docking
Docking was performed using the AutoDock Vina module incorporated into UCSF Chimera [40]. A grid box with coordinates of 51.3568, 154.894 and 29.7124, and x, y, and z dimensions of 23.009, 22.8623, and 26.0373 was used to define the location of the MAO-B binding site. Docked results generated were saved in pdbqt format. Conformations with best binding scores were selected and saved.

Molecular Dynamics Simulation
Six systems (five complexes bound with the inhibitors and unbound MAO-B (Apo)) were set up for molecular simulation using the GPU version incorporated in AMBER18 [41]. Forcefield FF14SB was used to define the systems. The ANTECHAMBER module was used to add atomic partial charges to ligands using the Restrained Electrostatic Potential (RESP) and the General Amber Force Field (GAFF) protocols [42].

Conclusions
The current study demonstrated a series of thiomethyl group containing chalcones as potent and selective class of MAO-B inhibitors. In conclusion, TM8 and TM1 were found to be reversible, selective, and competitive inhibitors of MAO-B with K i values of 0.0031 ± 0.0013 and 0.011 ± 0.001 µM, respectively. The isosteric replacement of the A ring methoxy-with a thiomethyl-group and the introduction of different substituents on the B ring showed good MAO-B inhibition. Moreover, TM8 and TM1 were found to be safe as analyzed by an in vitro toxicity study. Additionally, the prooxidant and antioxidant levels can be retained by TM8 and TM1. Our findings show that TM8 and TM1 are selective inhibitors of MAO-B and have potential therapeutic value for the treatment of neurological disorders.