Introduction

Of the heterogeneous and complex neurodegenerative disorders (NDDs) that largely affect the elderly, Alzheimer’s disease (AD) and Parkinson’s disease (PD) are the most prevalent and their pathogenesis have been attributed to a variety of genomic, epigenomic, and environmental factors (Van Bulck et al. 2019). Mounting evidence indicates that drugs targeting a single pathway cannot adequately address the multifactorial pathophysiologies of NDDs (Geldenhuys et al. 2011). Oxidative stress, mitochondrial dysfunction, and imbalances in the levels of enzymes that control the metabolism of biogenic amines may promote NDD progression (Barnham et al. 2004; Lin and Beal 2006). On the other hand, several molecular scaffolds have been designed to simultaneously target entities such as choline esterase (ChE), monoamine oxidases (MAOs), and β-site amyloid precursor protein cleaving enzyme 1 (β-secretase, BACE-1), to retard NDD progression (Zhang et al. 2019).

Piperazine is a heteromonocyclic, six-membered ring containing two secondary nitrogen atoms, and diazacycloalkane with a non-planar, flexible nature that can interact hydrophobically and by hydrogen bonding with target enzymes. Furthermore, piperazine is a privileged structure with acknowledged “drug-likeness” and balanced pharmacodynamic and pharmacokinetic properties (Rathi et al. 2016). Currently, more than 40 piperazine-containing drugs have been FDA-approved as antianginals, antidepressants, antiserotonergics, urologicals, anthelmintics, antineoplastic agents, nootropics, and tranquillizers (Brito et al. 2019). Furthermore, the substitution of one or both nitrogen atoms in the piperazine ring system with various structural motifs can result in significant MAO-A, MAO-B, and acetylcholinesterase (AChE) inhibitions (Pettersson et al. 2012; Kaya et al. 2017; Kumar et al. 2018; Özdemir et al. 2020; Sağlık et al. 2020; Jevtić et al. 2020; Modh et al. 2013; Sahin et al. 2018).

On the other hand, 1,3-diphenylprop-2-en-1-one is a chalcone that contains two phenyl rings separated by rotatable three-carbon units. This linker is an α, β-unsaturated ketone with trans-orientated olefinic linkage (Zhuang et al. 2017). More than 90% of the chalcones synthesized exhibit selective MAO-B inhibition (Chimenti et al. 2009; Guglielmi et al. 2020), and analogs of chalcones with furan, thiophene, indole, imidazole, or morpholine heterocyclic entities on the A ring of α, β-unsaturated ketones have been reported to act as competitive, selective, and reversible MAO-B inhibitors (Robinson et al. 2013; Mathew et al. 2016; Sasidharan et al. 2016, 2018; Mathew et al. 2019a). However, the effect of introducing the piperazine pharmacophore into the chalcone framework has not been explored in the context of MAO-B inhibition. Previously, new chalcones containing the piperazine or 2,5-dichlorothiophene moiety were synthesized and evaluated for antimicrobial activities (Tomar et al. 2007). Herein, we describe the synthesis of a series of piperazine-derived chalcones and inspected for their abilities to inhibit MAOs, ChEs, and BACE-1 in an effort to identify novel treatments for NDDs.

Materials and methods

Synthesis

Derivatives were synthesized using base-catalyzed Claisen-Schmidt condensation reaction between various aromatic aldehydes and 4′-piperazinoacetophenone (Tomar et al. 2007). Briefly, the synthesis was initiated by adding 0.01 M 4′-piperazinoacetophenone to an ethanol (20 ml)/40% of KOH (8 ml) mix and then adding 0.01 M of an aromatic benzaldehyde and stirring for 10–12 h (Scheme 1). Resulting solutions were poured into ice-cold water, and the precipitates were washed with water. The formation of the products was checked by thin layer chromatography with a solvent system of ethyl acetate: hexane (1:9). Recrystallization was done with ethanol.

Scheme 1
scheme 1

The synthetic route used

Full size image

Enzyme assays

Kynuramine (0.06 mM) and benzylamine (0.3 mM) were used as substrates to assay the activities of recombinant human MAO-A and MAO-B, respectively [Mathew et al. 2018]. AChE and butyrylcholinesterase (BChE) activities were assayed using Electrophorus electricus Type VI-S and Equine serum Type, respectively, in the presence of 0.5 mM acetylthiocholine iodide (ATCI) or 0.05 mM butyrylthiocholine iodide (BTCI), respectively, by adding 0.5 mM 5, 5′-dithiobis (2-nitrobenzoic acid) (DTNB) [Ellman et al. 1961; Baek et al. 2018a; Lee et al. 2019]. Enzymes and inhibitors were preincubated for 15 min before measuring inhibitory activities. BACE-1 activity was determined using a β-secretase (BACE-1) activity detection kit, which included the 7-methoxycoumarin-4-acetyl-[Asn670,Leu671]-amyloid β/A4 protein fragment 667-676-(2,4-dinitrophenyl)Lys-Arg-Arg amide trifluoroacetate as a substrate. The reaction was performed for 2 h at 37°C and the signals were measured using a fluorescence spectrometer (FS-2, Scinco, Seoul, Korea) with an excitation wavelength of 320 nm and an emission wavelength of 405 nm. Chemicals, enzymes, and BACE-1 activity detection kits were purchased from Sigma-Aldrich (St. Louis, MO, USA).

Enzyme inhibitory and kinetic studies

The inhibitory activities of the 11 compounds were initially measured at a concentration of 10 μM against MAO-A, MAO-B, AChE, BChE, or BACE-1. IC50 values for MAO-A and MAO-B by the compounds were determined first, and then, those for AChE, BChE, and BACE-1 by compounds with residual activities of < 50% were investigated. Kinetic studies were carried out on PC10 and PC11, which most potently inhibited MAO-B, at five substrate concentrations and three inhibitor concentrations, as previously described [Çeçen et al. 2020].

Inhibitor reversibility analysis

The reversibility of MAO-B inhibitions by PC10 and PC11 was assessed by dialysis after preincubating them at 0.15 μM with MAO-B for 30 min, as described previously [Baek et al. 2018b]. For comparison purposes, MAO-B was preincubated with 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 assessed by comparing the activities of dialyzed (AD) and undialyzed (AU) samples.

ADME prediction

ADME parameters including pharmacokinetic data and physical-chemical properties such as lipophilicity and water solubility of PC10 and PC11 were predicted in silico using free software available at http://www.swissadme.ch/ (Daina et al. 2017).

Molecular docking

The 3D-coordinates of MAO-A, MAO-B, and AChE crystals were collected from the Protein Data Bank by selecting 2Z5X, 2V5Z, and 4EY7 entries, respectively (Son 2008; Binda et al. 2007; Cheung et al. 2012). Protein Preparation Wizard tools were employed to optimize and minimize crystal structures (Schrödinger 2020c; Madhavi Sastry et al. 2013). To explore the chirality, ionization states, ring conformations, and tautomers of each input structure, PC4, PC10, and PC11 were treated by using the Ligprep tool (Schrödinger 2020b). Docking simulations were carried out using GLIDE (Friesner et al. 2004; Schrödinger 2020a), and the centers of mass of cognate ligands were used to generate the enclosing boxes. The standard precision docking protocol was used with default Force Field OPLS_2005 and detailed analysis of ligand binding affinities was performed by calculating binding free energies (ΔG) between protein and ligands using the Molecular Mechanics/Generalized Born Surface Area (MM-GBSA) method (Banks et al. 2005; Genheden and Ryde 2015). The ΔG values were calculated using: ΔGbind = ΔEMM + ΔGsolv + ΔGSA, where ΔEMM = minimized energy of the ligand-protein complex, ΔGsolv = solvation energy, and ΔGSA = surface area energy.

Results

Inhibitory activities against MAOs, ChEs, and BACE-1

All eleven compounds exhibited high inhibitory activities against MAO-B with residual activities of <50% at the concentration of 10 μM (Table 1). PC10 exhibited the greatest inhibitory activity against MAO-B (IC50 value = 0.65 μM), followed by PC11, which had an IC50 value of 0.71 μM. PC1 inhibited MAO-B least (IC50 = 7.62 μM), and the other 8 compounds had IC50 values ranging from 1.09 to 3.65 μM. A comparison of the IC50 value of PC1 with those of PC10 and PC11 indicated that the presence of a –F or –CF3 group instead of –H increased MAO-B inhibitory activity. Other substituents such as –OCH3 of PC3, –Cl of PC8, and –OH of PC2 moderately enhanced MAO-B inhibition. However, all compounds much less effectively inhibited MAO-A at 10 μM and achieved residual activities of > 65% (Table 1). PC3 had the lowest IC50 value of 27.9 μM. Regarding selectivity index (SI), PC10 and PC11 had the highest values of 48.3 and 49.2, respectively, for MAO-B over MAO-A.

Table 1 Inhibitions of recombinant human MAO-A, MAO-B, AChE, BChE, and BACE-1 by piperazine-substituted chalconesa
Full size table

All compounds weakly to moderately inhibited AChE and BChE by < 50% at 10 μM, except PC4, which inhibited AChE by 56.6% (IC50 = 8.77 μM) (Table 1). PC10 and PC11 had IC50 values of 28.0 and 26.3 μM, respectively, for AChE, and showed moderate inhibitory activities against BChE with IC50 values of 36.4 and 36.2 μM, respectively. Interestingly, some of the compounds effectively inhibited BACE-1 (Table 1). PC3 effectively inhibited BACE-1, and PC10 and PC11 showed moderate BACE-1 inhibitory activities with IC50 values of 14.9 and 15.3 μM, respectively; these two compounds contained fluoro and trifluoromethyl pharmacophores, respectively, on a piperazine-substituted chalcone framework.

MAO-B inhibition and its kinetics

MAO-B inhibitions by PC10 and PC11 were subjected to kinetic analysis. Lineweaver-Burk and secondary plots showed that PC10 and PC11 competitively inhibited MAO-B (Fig. 1a and c), with Ki values of 0.63 ± 0.13 and 0.53 ± 0.068 μM, respectively (Fig. 1b and d). These experimental observations showed that PC10 and PC11 are competitive and selective inhibitors of MAO-B.

Fig. 1
figure 1

Lineweaver-Burk plots for MAO-B inhibition by PC10 (a) or PC11 (c), and respective secondary plots (b and d) of slopes vs. inhibitor concentrations

Full size image

Reversibility studies of MAO-B inhibition

MAO-B inhibitions by PC10 and PC11 were also subjected to reversibility studies. The results obtained showed that inhibition of MAO-B by PC10 was recovered after dialysis from 32.1 (AU value) to 81.2% (AD value), and that inhibition by PC11 recovered from 32.6 (AU) to 68.9% (AD). These values were close to those of the reversible reference lazabemide (from 28.3 to 86.9%). On the other hand, inhibition of MAO-B by the irreversible reference pargyline was not recovered by dialysis (from 31.8 to 32.7%) (Fig. 2). The above data revealed that MAO-B inhibitions by PC10 and PC11 were recovered to reversible reference values and showed that PC10 and PC11 reversibly inhibit MAO-B.

Fig. 2
figure 2

Recoveries of MAO-B inhibitions by PC10 and PC11 using dialysis experiments

Full size image

ADME prediction

In silico ADME prediction revealed that PC10 and PC11 fully complied with the limits of Lipinski’s rule and supported their oral use and drug-likenesses (Table 2). In addition, the boiled-egg and bioavailability radar plots generated by the SwissADME tool also provided two clear ADME characteristics (Fig. 3). Boiled-egg pictures represent passive absorption in the gastrointestinal (GI) tract (shown as the white area) and the ability to cross the blood-brain barrier (BBB; yellow area). These plots showed that PC10 and PC11 were located in the yellow area, suggesting ability to cross the BBB, and indicated that both would be passively absorbed in the GI system (Fig. 3a and b). Bioavailability radar also placed the drug-likeness representation of PC11 within the pink area, which is an indicator of optimal physicochemical properties, such as size, polarity, solubility, lipophilicity, saturation, and suitability for oral administration. In this respect, PC10 showed a slight deviation in physicochemical property of saturation (Fig. 3c and d).

Table 2 ADME predictions for PC10 and PC11
Full size table
Fig. 3
figure 3

Representations of the boiled-egg graphs (a and b) and bioavailability radar plots (c and d) for PC10 and PC11 produced using the SwissADME web-tool

Full size image

Molecular docking studies

Molecular docking simulations and binding free energies calculations were carried out to investigate molecular interactions between PC4, PC10, or PC11 and the binding sites of MAOs and AChE. PC10 and PC11 had docking scores of −5.73 and −4.81 kcal/mol, respectively, toward MAO-A, and of −7.29 and −7.11 kcal/mol, respectively, toward MAO-B (Table 3). In addition, the MM-GBSA values of PC10 and PC11 were −28.77 and −52.60 kcal/mol, respectively, toward MAO-A, and −64.19 and 66.06 kcal/mol, respectively, toward MAO-B (Table 3). Regarding AChE inhibition, although docking simulation returned comparable scores for the three compounds, PC4 had the lowest binding free energy (i.e., −55.81, −39.82, and −52.18 kcal/mol for PC4, PC10, and PC11, respectively) (Table 3).

Table 3 Docking scores and MM-GBSA values for interaction between PC4, PC10, or PC11 and MAO-A, MAO-B, or AChE
Full size table

The para-methyl and para-fluorine rings of PC4 and PC10 faced FAD and interacted through π–π contacts with Y407 of MAO-A (Fig. 4a). Conversely, although slightly larger, the −CF3 substituent of PC11 did not allow its styrene moiety to face FAD, which resulted in the phenyl ring of chalcone establishing π–π interactions with F208 of MAO-A. All three compounds assumed similar binding poses within the MAO-B binding pocket (Fig. 4b), which was ascribed to the existence of two π–π interactions between chalcone aromatic rings and Y398 or Y326 of MAO-B selective residues. The styrene portion of PC4 established a π–π interaction with Y341 of AChE and the carbonyl group of its chalcone moiety formed a hydrogen bond with F295 of AChE (Fig. 4c). On the other hand, the chalcone moieties of PC10 and PC11 formed π–π interactions with W286 and hydrogen bonds with the main chain of F295.

Fig. 4
figure 4

Molecular dockings with the binding pockets of MAO-A (a), MAO-B (b), or AChE (c) by PC4 (a, yellow sticks), PC10 (b, green sticks), or PC11 (c, cyan sticks), respectively. Green lines and red arrows indicate π–π contacts and hydrogen bonds, respectively. Water molecules are depicted as cyan spheres. Docking score values for PC4, PC10, or PC11 with MAO-A were −5.845, −5.730, and −4.811 kcal/mol, respectively; with MAO-B were −7.293, −7.285, and −7.113 kcal/mol, respectively; and with AChE were −6.099, −5.584, and −6.681 kcal/mol, respectively

Full size image

Discussion

Chalcones are considered versatile scaffolds for many of the CNS-related agents, such as anti-depressants, anxiolytics, β-amyloid plaque imaging agents, adenosine receptor antagonists, and MAO-B and AChE inhibitors (Mathew et al. 2019b). The three rotatable bonds available in the Michael acceptor between the two phenyl systems of chalcones provide different mode interactions in the inhibitor binding cavities of the enzymes, which are highly dependent on the natures and bulkinesses of groups bearing on the A and B chalcone rings (Matos et al. 2015).

Recent studies have reported that the presence of electron-donating groups, such as methyl, methoxy, ethyl, dimethylamino, and ethyl acetohydroxamate, at the para position of the phenyl B ring of chalcones confers greater MAO-B inhibition than MAO-A inhibition. Lipophilic halogen atoms (fluorine, chlorine, and bromine) at the same position also resulted in outstanding MAO-B inhibition (Morales-Camilo et al. 2015; Reeta et al. 2019; Shalaby et al. 2019). In addition, the presence of an aliphatic or methyl-containing amino group or a nitrogen-derived pharmacophore in chalcones is required for AChE inhibition (Liu et al. 2016; Xiao et al. 2017; Bai et al. 2019). In this respect, the present design strategy explores the effects of the presence of both pharmacophores in the chalcone framework (Fig. 5) by introducing a piperazine nucleus at the para position of the phenyl A ring and various electron-donating or electron-withdrawing substituents on the B ring of the chalcone scaffold. Recently, Sasidharan et al., reported that morpholine-bearing chalcones exhibited dual-acting inhibitory activities, i.e., selective MAO-B inhibition with moderate AChE inhibition (Sasidharan et al. 2021).

Fig. 5
figure 5

Design strategy used to produce piperazine-based multi-target directed ligands (MTDLs)

Full size image

In the present study, the unsubstituted piperazine-based chalcone (PC1) exhibited moderate MAO-B inhibition with an IC50 value of 7.62 μM but high residual activities for MAO-A, AChE, BChE, and BACE-1 (96.6%, 72.9%, 96.0%, and 90.6%, respectively) at 10 μM. The introductions of small groups on the B ring of the phenyl system slightly impacted activity ratio toward multi-targets, which emphasizes the importance of substituents on the B ring. Some interesting structure-activity relationships (SARs) were derived, as depicted in Fig. 6.

Fig. 6
figure 6

SAR analyses of the 11 piperazine-based chalcones produced

Full size image

The SAR studies revealed that all piperazine-substituted chalcones inhibited MAO-B substantially or moderately better than MAO-A. The presence of electron-withdrawing groups like trifluoromethyl or fluorine provided good MAO-B inhibition and the presence of a methyl group (electron-donating) on the para position of the phenyl B ring provided optimal activity as exemplified by PC4, which had IC50 values for MAO-B and AChE of 2.72 and 8.77 μM, respectively, and inhibited BACE-1 by 44.5% at 10 μM. Whereas the introduction of electron-donating groups like methyl (PC4), hydroxyl (PC2), or methoxy (PC3) groups resulted in moderate AChE inhibition (residual activities of 43.4%, 56.8%, and 58.8%, respectively, at 10 μM). Moreover, these electron-donating groups showed moderately selective inhibition of MAO-B. Regarding the effects of halogen substitution at phenyl ring on MAO-B inhibition, fluorine (PC10) had a greater effect than chlorine (PC8) or bromine (PC9). The high electronegativity of fluorine substituted phenyl in various molecular frameworks has been recently reported to increase hydrophobic interactions with Tyr398 and Tyr435 markedly in the active site of MAO-B (Mathew et al. 2020).

MAO-B, AChE, and BACE-1 are important enzyme targets and can produce neurotoxic free radical by-products, degrade acetylcholine, and generate amyloid β (Aβ), and thus, the development of highly selective ligands that target these enzymes is of considerable interest to those involved in the development of drugs for AD (Moussa-Pacha et al. 2019; Benny and Thomas 2019). Recently, many researchers have used ligand or structure-based drug strategies to design dual-acting MAO-B/AChE inhibitors (Mathew 2020). However, this is the first study to focus on the inhibitory profiles of piperazine containing chalcone-based compounds on MAO-A, MAO-B, AChE, BChE, and BACE-1. Our findings show that the lead molecules PC4, PC10, and PC11 selectively inhibit MAO-B at the submicromolar level and moderately inhibit AChE and BACE-1.

Reversible MAO-B inhibitors have shorter action durations than irreversible inhibitors because they dissociate from targets (Tipton 2018). Many new reversible MAO-B inhibitors have been developed using various structural scaffolds, such as chromones, coumarins, chalcones, phenyloxazolidinones, and pyrazolines (Carradori and Silvestri 2015; Mathew et al. 2017; Mathew 2020). Our reversibility studies on PC10 or PC11 inhibition of MAO-B indicate that the reversible natures of their interactions cause minimal target disruption and provide improved ADME profiles.

Computational results of interactions between PC4, PC10, or PC11 and MAO-A, MAO-B, or AChE targets provided satisfactory explanations of experimental data and agreed well with experimental IC50 values. Furthermore, they showed that π–π interaction with Y326 of MAO-B (a selective residue), which is changed to I335 in MAO-A (Mangiatordi et al. 2017), is a prerequisite for interaction in its binding pocket, as previously reported (Oh et al. 2020).

Conclusion

We report the results of an investigation of the neuro-related, multi-enzyme targeting profiles of piperazine-bearing α,β-unsaturated ketones. The abilities of MAOs and ChEs inhibitions by the compounds were found to depend on the natures of substituents at the para position of the B phenyl ring of the chalcone framework. All eleven compounds inhibited MAO-B more than MAO-A, and compounds PC4, PC10, and PC11 potentially inhibited MAO-B with IC50 values of 2.72, 0.65, and 0.71 μM, respectively, and moderately inhibited AChE with IC50 values of 8.77, 28.0, and 26.3 μM, respectively. Furthermore, these experimental results were also supported by molecular docking studies. In addition, PC4, PC10, and PC11 moderately inhibited BACE-1 with IC50 values of 15.5, 14.9, and 15.3 μM, respectively, which further supports their potential use for the development of novel drugs against various neurological disorders.