Overcoming undesirable hERG aﬀinity by incorporating fluorine atoms: A case of MAO-B inhibitors derived from 1 H-pyrrolo-[3,2-c]quinolines

The incorporation of the ﬂ uorine motif is a strategy widely applied in drug design for modulating the activity, physicochemical parameters, and metabolic stability of chemical compounds. In this study, we attempted to reduce the af ﬁ nity for ether- (cid:3) a-go-go-related gene ( h ERG) channel by introducing ﬂ uorine atoms in a group of 1 H -pyrrolo[3,2- c ]quinolines that are capable of inhibiting monoamine oxidase type B (MAO-B). A series of structural modi ﬁ cations guided by in vitro evaluation of MAO-B inhibition and antitargeting for h ERG channels were performed, which led to the identi ﬁ cation of 1-(3-chlorobenzyl)-4-(4,4-di ﬂ uoropiperidin-1-yl)-1 H -pyrrolo[3,2- c ]quinoline ( 26 ). Compound 26 acted as a reversible MAO-B inhibitor exhibiting selectivity over 45 targets, enzymes, transporters, and ion channels, and showed potent glioprotective properties in cultured astrocytes. In addition, the compound demonstrated good metabolic stability in rat liver microsomes assay, a favorable safety pro ﬁ le, and brain permeability. It also displayed procognitive effects in the novel object recognition test in rats and antidepressant-like activity in forced swim test in mice. The ﬁ ndings of the study suggest that reversible MAO-B inhibitors can have potential therapeutic applications in Alzheimer's disease.


Introduction
Alzheimer's disease (AD), a severe neurodegenerative disorder, which constitutes the most common form of dementia, is becoming more common, posing a significant burden on public health. The currently available pharmacotherapy for AD mainly involves symptomatic treatment. Although the recent approval of Aducanumab, a monoclonal antibody directed against amyloid beta (Ab), seemed promising for the development of disease-modifying strategies, such an approach requires further trials to verify its expected clinical benefits in AD. Therefore, there remains an urgent need for novel AD treatments.
Monoamine oxidase type B (MAO-B) is a flavin adenine dinucleotide (FAD)-containing enzyme located in the mitochondrial membrane of serotonin and histaminic neurons [1] and glial cells [2,3]. It catalyzes the oxidative deamination of neurotransmitters, such as dopamine [4,5], which is accompanied by the generation of reactive oxygen species causing toxicity to neuronal and glial cells [6].
The beneficial effects of MAO-B blockade in the early symptomatic treatment of Parkinson's disease have been well documented. MAO-B inhibitors may be used alone as monotherapy or as add-on drugs (e.g., to levodopa). Nevertheless, these compounds gained interest as a potential approach to treating the cognitive decline associated with AD, considering the increased MAO-B levels in the brains of AD patients [7,8]. Indeed, the elevated level of MAO-B has been reported as an early event of AD which persists during the disease progression [9,10]. Additionally, an up-regulated MAO-B activity coincides with the presence of reactive astrocytes, during neuroinflammatory process in AD [11].
Clinical trials revealed that short-term treatment with selegiline, an irreversible MAO-B inhibitor, might reverse cognitive deficits in AD patients [12], whereas long-term administration of this drug had no significant effects [13]. Recent findings linking the activity of MAO-B and the expression of g-aminobutyric acid (GABA) shed new light on the type of MAO-B inhibition and its potential impact on memory decline in AD. Pharmacological inhibition of MAO-B with irreversible inhibitors, which involves covalent modification of the active site of the enzyme, results in the destruction of the enzyme and the activation of alternative mechanisms to compensate for the level of GABA in reactive astrocytes [14]. On the other hand, reversible inhibitors (e.g., safinamide), which compete with the substrate at the binding site of MAO-B, do not damage the enzyme and thus do not induce unfavorable compensatory mechanisms [14].
In our recent studies on the series of 1H-pyrrolo [3,2-c]quinoline derivatives characterized by decreased basicity [15] and expected serotonin type 6 receptor (5-HT 6 R) antagonistic properties [16e18], a new MAO-B inhibitor, compound II, was identified. This compound exhibited a decreased affinity for 5-HT 6 R which resulted from the replacement of the basic nitrogen in the pyrrolidine fragment with an oxygen atom (I, II vs CPPQ, Fig. 1) [16]. The affinity for 5-HT 6 R (II vs CPPQ) was further reduced when the N 1 -sulfonyl moiety was replaced with a methylene bridge.
These results led to the search for MAO-B inhibitors among 1Hpyrrolo [3,2-c]quinolines and prompted us to investigate the structureeactivity relationship of these compounds. In addition, we optimized the synthesis procedure for producing the PQ core using less toxic reagents. Structural modifications involved electron-donating and electron-withdrawing substituents at benzyl fragment, as well as alicyclic and aliphatic amines, aminoethers, and aminoalcohols at position 4, of the 1H-pyrrolo- [3,2-c] quinoline (PQ) core ( Fig. 1).
Though some of the newly obtained derivatives inhibited MAO-B activity at a nanomolar or subnanomolar concentration, they exhibited high affinity for the human ether-a-go-go-related gene (hERG) channel, which is responsible for the prolongation of the QT interval. Therefore, we performed off-target driven optimization to gain new structural insights into MAO-B inhibitors that do not exhibit affinity to hERG channel. Furthermore, we assessed the pharmacokinetic properties of the most promising derivative and its glioprotective effect in cultured astrocytes exposed to cytotoxic doxorubicin (DOX). Finally, we carried out an in vivo evaluation of selected compounds to verify whether reversible MAO-B inhibitors derived from the PQ core could be promising compounds for the treatment of AD and exhibit procognitive properties and antidepressant-like activity.

Chemistry
The previously reported 1H-pyrrolo [3,2-c]quinoline core 7 [16] was synthesized using a procedure involving the aza-BayliseHillman reaction, N-allylation, and ring-closing metathesis (RCM) as the key steps. Although the optimized flow-chemistry approach used for RCM was fast and more environmentally friendly [19], the first two steps required the use of harmful reagents (methyl acrylate and allyl bromide) [16,17]. In order to replace this procedure with more sustainable methods, we applied the Leuckart-Wallach approach to convert commercially available 2-nitrobenzaldehyde 1 into respective formamide 2, followed by the formation of isocyanide 3 after treatment with POCl 3 (Scheme 1) [20].
The resulting product 3 was subsequently heated with methyl propiolate in the presence of silver carbonate, which led to the formation of a nitro-derivative of methyl 2-phenyl-1H-pyrrole-3carboxylate 4. This allowed the synthesis of 2-aryl-1H-pyrrole carboxylate 4 in an overall yield of 30%, which is significantly higher than the yield of 21% obtained with the initial protocol (Table 1-SI) [21]. In addition, this approach enabled the synthesis of 2-aryl-1Hpyrrole-3-carboxylate 4 in three steps (one step less than previously), with the use of less toxic reagents compared to the RCMbased method (Scheme 1-SI) [21]. The subsequent steps proceeded following the reported synthetic routes. Reduction of the nitro group of 4 yielded aromatic amine derivative 5, which was cyclized to lactam 6, and then underwent a chlorodehydroxylation with POCl 3 to obtain 1H-pyrrolo[3,2-c]quinoline 7. [16].
Heating of compound 7 with respective benzyl bromides in the presence of phosphazene base P1-t-Bu-tris(tetramethylene) (BTPP) yielded benzyl derivatives 8ae8h (Scheme 2). Subsequent coupling with primary amines occurred under BuchwaldeHartwig N-arylation conditions, while reactions with secondary amines were carried out under prolonged microwave heating in acetonitrile in the presence of triethylamine (TEA).

Structureeactivity relationship studies
Molecular docking studies of representative library derivatives were carried out on five MAO-B crystals structures containing a significant amount of water molecules (W1eW6; Fig. 1-SI and 2-SI). Surprisingly, no ligandereceptor complexes were identified in the analysis, which indicated that the conformational state of the enzyme, and thus the position of water molecules, may vary depending on the MAO-B crystals. Therefore, we performed a fully flexible docking procedure (induced fit-docking (IFD)), in which the conformational change occurred in the enzyme active site, FAD, inhibitor, and W1eW6 water molecules. Based on the best conformations determined in IFD, a series of dockings were performed for the selected compounds.
A comparative analysis of the binding modes of the initially identified analogs I and II (Fig. 2) revealed that the 2aminotetrahydrofuranyl fragment was localized within the hydrophobic crevice formed by the FAD tricyclic system and Y398, Y326, and Y435 amino acid residues. The analysis showed a slight rotation of this group within both derivatives and the formation of a hydrogen bond with Y326. Six water molecules (W1eW6) were found to be conserved ( Fig. 2-SI) and seem to support the orientation of noncovalent inhibitors within the active site of the enzyme (Fig. 2)  positioned in the center of the catalytic pocket, forming a hydrogen bond with water molecule W3 and taking part in hydrophobic interactions with F168 and Y326. The 3-chlorobenzenesulfonyl and 3-chlorobenzyl fragments were located near the B entrance channel in both derivatives. The rotation of these fragments toward the channel was blocked by water molecules W5 and W6, which made them point toward the hydrophobic cavity formed by amino acids L167, I316, L164, and F168. In addition, the Cl substituent present at position 3 stabilized the orientation of the 3-chlorobenzyl fragment by forming a weak halogen bond with the carbonyl oxygen of L164. The 3-chlorobenzyl derivative II was characterized by higher inhibitory activity compared to its 3-chlorobenzenesulfonyl analog, which might be due to the limited rotation of the latter in the rigid binding pocket of MAO-B.
In vitro studies revealed that the newly synthesized compounds 9e33 had moderate-to-high inhibitory effect on MAO-B, with their IC 50 values ranging from 0.7 to 289 nM, and no activity for MAO-A isoform (2e5% inhibition at 1 mM), as determined using the fluorometric method ( Table 1).
The expansion of the tetrahydrofuranyl (THF) ring of compound II to the six-membered tetrahydropyranyl moiety contributed to maintaining its inhibitory activity on MAO-B and high selectivity over 5-HT 6 R (II vs 9). Further separation of the THF ring from the exocyclic amino group with the use of a methylene bridge revealed the importance of the distance between nitrogen and oxygen atoms in the aminoether fragment (11 and 12): attachment of the methylene spacer at C 2 position of THF caused a significant decrease in the MAO-B inhibitory activity (11 vs 9), while a shift of the spacer to Table 1 Influence of the aminoether fragment on the MAO-B and MAO-A inhibitory activity, and affinity for 5-HT 6 R and hERG channel. C 3 position was well-tolerated and did not affect the inhibitory potential (12 vs 9).
On the other hand, all aminoether derivatives 13e17, containing an endocyclic nitrogen atom and an exocyclic oxygen atom, displayed potent MAO-B inhibitory effect. In particular, derivatives 14 and 15 bearing 4-methoxy piperidine showed inhibitory activity at subnanomolar concentrations.
Because several structurally and functionally unrelated central nervous system drugs can block the hERG channel, a potassium channel responsible for the prolongation of QT interval in electrocardiogram, we evaluated the affinity of selected compounds for this channel (Table 1) [22]. We found that compounds 10, 12, 14, and 15 showed a high affinity for hERG at a concentration of 1 mM (72e94% inhibition of dofetilide binding at 1 mM) in the screening Table 2 Influence of the spiroaminoether/aminoalcohol fragment on the MAO-B, and MAO-A inhibition activity, and affinity for 5-HT 6 R and hERG channel.  procedure. Therefore, we included a screen for hERG activity, as a key antitarget, in our lead-optimization workflow.
We further hypothesized that activity at hERG might be diminished by the removal of Lewis base oxygen atom. Nevertheless, the introduction of a piperidine moiety at position 4 of the pyrroloquinoline scaffold (25) did not cause a considerable decrease in the affinity for this antitarget (Table 3).
It is worth noting that the introduction of fluorine atoms in the structure of the molecule, which results in a decrease in the pK a value, is one of the strategies applied for detuning the hERG activity and improving metabolic stability of compounds [23]. Therefore, we introduced a gem-difluoro substituent at position 4 of piperidine fragment to obtain 26. This modification allowed maintaining potent MAO-B inhibitory activity as well as safe hERG affinity (26 vs 25). In silico docking studies confirmed that derivative 26 showed similarities to one of the most potent compounds from the series bearing 4-methoxypiperidine moiety (14; Fig. 3A) in terms of the location and type of interaction of the 4,4-difluoropiperidine fragment.
In parallel, the in vitro metabolic stability of compounds 25 and 26 was determined using the rat liver microsomes (RLM) assay. The results showed that compound 25 displayed a high intrinsic clearance (104 ml/min/kg), whereas its analog 26 showed lower clearance (29 ml/min/kg). This suggested that the gem-difluoro substituent had a positive effect on the biological properties of compounds.
Following the identification of 4,4-difluoropiperidin-1-yl as the optimal amine fragment, we further investigated the influence of the position of the substituent in the benzyl fragment on the MAO-B inhibitory activity (Table 4).
Using a computational protocol involving molecular dynamic simulations, ab initio docking (quantum polarized ligand docking, QPLD), and energy calculations applying the Molecular Mechanics Generalized Born Surface Area (MM-GBSA) method, we first estimated the energy gain of the halogenated derivatives at positions 2, 3, and 4 relative to the unsubstituted derivative 27. The results showed that the derivatives substituted at position 3 gained the highest stabilization energy of the complex, whereas 2-and 4substituted derivatives formed less stable complexes due to steric hindrance (Table 4). It was also found that the complex was additionally stabilized by the halogen bond formed between the 3-Cl substituent and the carbonyl oxygen of L164 (Fig. 3B).
In vitro studies revealed that the presence of substituent at position 3 favored the inhibitory activity (position 3 > position 2 > position 4). On the other hand, removal of the initially introduced chlorine atom (27 vs 26) as well as its replacement with fluorine (30) or electron-donating methoxy group (31 vs 26) was unfavorable. Thus, the applied modifications confirmed the importance of the chlorine atom at position 3 of the aromatic ring.  Selected compounds (IC 50 < 120 nM) bearing the 4,4difluoropiperidin-1-yl fragment displayed good metabolic stability in the RLM assay (Table 5).

Assessment of the MAO-B inhibitory activity of compound 26
Kinetic studies using in vitro methods were performed to assess the inhibitory effect of compound 26 on MAO-B. In the first step, the reversibility of compound 26 was evaluated. Rasagiline and safinamide, which displayed irreversible and reversible MAO-B inhibitory activity, respectively, were used as references. To assess the inhibitor's activity tested compounds were preincubated with the enzyme and low concentration of the substrate for 30 min, followed by the 100-fold increase of substrate concentration. As can be seen on the plots obtained for rasagiline, an irreversible inhibitor, its preincubation with the enzyme resulted in the irreversible decrease of fluorescent product (resorufin) formation in MAO-B/Horseradish peroxidase (HRP)-coupled reaction (Fig. 4). In contrast, safinamide and 26 time-dependently increased MAO-B enzymatic activity, and thus behaved as reversible inhibitors (Fig. 4).
Compound 26 was then tested at three concentrations corresponding to its IC 20 , IC 50 , and IC 80 values (4, 17, and 66 nM, respectively) to evaluate its reversible inhibition modality (Fig. 5). The results obtained from the kinetic experiment were used for constructing MichaeliseMenten curves related to the substrate concentration and determining the respective reaction rate (Fig. 5A). Transformation of the obtained data using the LineweavereBurk equation allowed generating a double-reciprocal plot (Fig. 5B). As could be seen on the LineweavereBurk plot, lines converge to the left of the y-axis and above the x-axis, which suggests a mixed mode of inhibition. Therefore, we assumed that compound 26 can bind unequally to the free enzyme and enzymeesubstrate complex, displaying a higher affinity for the former. It should be noted that in our experimental conditions, safinamide, a reversible MAO-B inhibitor used as a reference, also displayed a mixed mode of inhibition ( Fig. 5C and D).

Selectivity profiling
Because early identification of the compound's affinity for offtargets might reduce safety-related attrition in further development, selectivity profiling was performed for compound 26. We found that the lead compound showed significant selectivity for over 45 receptors (Table 6), enzymes, transporters, and ion channels (Table 7).
Interestingly, compound 26 displayed an affinity for sodium channels in the nonselective [ 3 H]-batrachotoxin binding (83.7% inhibition at 10 mM) ( Table 7). Since blockade of sodium channels may cause inhibition of glutamate release from stimulated nerve terminals [24], this mechanism is regarded as a valid strategy of reducing overactive glutamate transmission. It may thus constitute an important added value to the MAO-B inhibitory activity of compound 26 [25]. Compound 26 did not inhibit the activity of sodium Nav1.5 channels (7.7% at 10 mM), which are one of the key antitargets involved in the pathogenesis of cardiomyopathy (Table 7).
Moreover, compound 26 did not activate serotonin 5-HT 2B Rs (À15% at 10 mM) (Table 6), and the preliminary result from the dofetilide binding assay showed that it displayed weak interaction with hERG channels (57% in electrophysiological QPatch experiment at 10 mM) ( Table 7). This indicates that compound 26 does not induce cardiotoxic effects associated with these targets, such as valvulopathy and cardiac arrhythmia [26].

Preliminary safety profile assessment and evaluation of ADME/ Tox properties
Modulation of cytochrome P450 (CYP450) can result in undesirable drugedrug interactions, which is an important safety concern in the drug development process. Because some MAO-B inhibitors have been recently reported to influence the activity of CYP450 [27], we evaluated compound 26 for its inhibitory effect on these enzymes. Compound 26 decreased the activity of CYP2C9 and CYP3A4 (21% and 59% respectively), but had no effect on CYP2D6 (92%).
Because drug-induced liver injury is one of the most serious concerns associated with drug withdrawal, compound 26 was subsequently tested in the human liver cancer cell line (HepG2) to assess its hepatotoxic effect. The MTS (3-(4,5-dimethylthiazol-2yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium inner salt) assay revealed that compound 26 did not affect the viability of the tested cells after exposure for 24 h (Fig. 3-SI). Additionally, the MTT (3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide) assay performed with an aim of assessing mitochondrial metabolism in C8-D1A astrocytes indicated that the compound 26 did not induce significant cytotoxicity at concentrations up to 5 mM. Finally, the AMES (Salmonella typhimurium reverse mutation) assay carried out for assessing mutations in the genes engaged in the histidine synthesis revealed that compound 26 did not show mutagenic effect at any tested concentration (1 or 10 mM) (Fig. 4-SI).

Glioprotective effects
Due to the uptake and degradation of Ab, glial cells, particularly astrocytes, have been shown to play a protective role in the early stage of AD. However, as the disease progresses, the astrocyte clearance of this protein/peptide decreases [28]. The resulting accumulation of Ab stimulates astrocytes to produce proinflammatory agents [29].
We examined the glioprotective effects of compound 26 in a model of cultured astrocytes (C8-D1A), exposed to DOX, a cytotoxic agent. DOX-induced cell death occurs as a result of alterations in synaptic plasticity, apoptosis, and lipid peroxidation. The MTT assay showed that compound 26 protected C8-D1A astrocytes at 0.25 mM against DOX-induced cytotoxicity (Fig. 6). On the other hand, neither selegiline nor safinamide used as a reference compounds displayed a significant glioprotective effect.

Pharmacokinetics evaluation
To investigate the preliminary pharmacokinetic profile of compound 26, we determined its concentrations in plasma (Fig. 7a) and brain (Fig. 7b) in Sprague Dawley rats at various time points following oral and intravenous administration (3 mg/kg).
A direct comparison of the area-under-the-curve values determined after both oral and intravenous administration indicated that compound 26 displayed excellent bioavailability of 94.9%. Further analysis of brain tissue showed that it crossed the bloodebrain barrier and reached maximal concentration (C max ) of 110.7 ng/ml after 2 h of oral administration. Its brain-to-plasma concentration ratio after intravenous and intragastric administration was 1.03 and 1.43, respectively. Compound 26 was eliminated from the body with a half-life of 148 and 188 min after intravenous and intragastric administration, respectively.

Effects on scopolamine-induced cognitive deficit
As cognitive impairment in AD is accompanied by decreased cholinergic neurotransmission, we assessed the potential procognitive effects of compound 26 in rats treated with anticholinergic scopolamine to induce cognitive deficits in the novel object recognition (NOR) test [30]. The results showed that compound 26 administered orally 2 h before the acquisition trial prevented scopolamine-induced short-term memory deficits (F(3,35) ¼ 15.7, p < 0.0001; Fig. 8).

Antidepressant properties
Since depression is a common comorbid condition in AD, we finally assessed the potential antidepressant-like activity of compound 26 in mice by performing forced swim test (FST). Of note, selegiline produced antidepressant-like activity in the in vivo preclinical models [31,32], and showed significant antidepressant effect in treatment-resistant depression [33]. Additionally, clinical trials confirmed efficacy of safinamide in treating depressive symptoms [34e36] and improving of the mood [35] in Parkinson's disease (PD) patients. Likewise, concomitant use of safinamide with antidepressant drugs (e.g., selective serotonin reuptake inhibitors) in PD patients was well-tolerated and such treatment did not evoke serotonin syndrome [36].
The results revealed that the compound (0.312e2.5 mg/kg) exhibited antidepressant-like properties, at one dose of 0.625 mg/ kg shortening the immobility time by 36.5% (Fig. 9). Of note, the potency of 26 administered at a dose of 0.625 mg/kg was similar to that of the antidepressant drug S-citalopram achieved with a twotime higher dose (1.25 mg/kg). Compound 26 showed a U-shaped doseeresponse curve, whereas S-citalopram dose-dependently shortened immobility time.
Active doses of the investigated compound 26 and the reference drug had no effect on the spontaneous locomotor activity measured over the observation period in the FST (i.e., from 2 to 6 min). This suggests that the antidepressant effects of these compounds are specific (data not shown).

Conclusions
Because of the well-known effects of fluorine on the physicochemical properties and biological activity of compounds, this element has been widely incorporated in chemical structures in the drug optimization process. In this study, we performed a systematic structureeactivity relationship analysis and identified compound 26 (1-(3-chlorobenzyl)-4-(4,4-difluoropiperidin-1-yl)-1H-pyrrolo [3,2-c]quinoline), a potent reversible MAO-B inhibitor. Most importantly, this compound was obtained using less toxic agents, by applying a new protocol for the synthesis of pyrroloquinoline core. The obtained compound 26 displayed a low affinity for the hERG channel as well as good metabolic stability, which could be attributed to the introduction of fluorine atoms. Moreover, the compound showed selectivity for over 45 targets, a favorable safety profile, and brain permeability. The derivative 26 also demonstrated glioprotective effects against DOX in a model of cultured astrocytes. In in vivo evaluation, compound 26 reversed scopolamine-induced cognitive deficits in the NOR test in rats and displayed antidepressant-like activity in FST in mice. Taken together, these results suggest that fluorination strategy can be successfully applied in medicinal chemistry for the development of selective MAO-B inhibitors devoid of hERG channel activity.

General methods
The synthesis was conducted at room temperature, unless indicated otherwise. Organic solvents (from Sigma-Aldrich and Chempur) were of reagent grade and were used without purification. All reagents (Sigma-Aldrich, Fluorochem and TCI) were of the highest purity.   Column chromatography was performed on silica gel Merck 60 (70e230 mesh ASTM).
UPLC and MS were carried out on a system consisting of a Waters Acquity UPLC coupled to a Waters TQD mass spectrometer. All the analyses were carried out using an Acquity UPLC BEH C18 100 Â 2.1 mm 2 column at 40 C. A flow rate of 0.3 ml/min and a gradient of (0e100)% B over 10 min was used: eluent A, water/0.1% HCOOH; eluent B, acetonitrile/0.1% HCOOH. Retention times, t R , were given in minutes. The UPLC/MS purity of all the test compounds and key intermediates was determined to be >95%.
High-resolution MS measurements were performed on a Bruker Impact II mass spectrometer (Bruker Corporation, Billerica, USA). Electrospray ionization (ESI) was used in the positive ion mode. Mass accuracy was within 2 ppm error in full-scan mode. The optimized MS parameters were the following: ion spray voltage 4 kV; capillary temperaturę 240 C, dry gas flow rate 4 l/min. Highpurity nitrogen as the nebulizing gas was used. Samples of 50 mM concentration were prepared from tested compounds using an eluent of acetonitrile/water 80:20 (v/v)/1% HCOOH. 1 H NMR and 13 C NMR spectra were recorded using JEOL JNM-ECZR 500 RS1 (ECZR version) at 500 and 126 MHz, respectively and are reported in ppm using deuterated solvent for calibration (CDCl 3 , methanol-d 4 or dmso-d 6 ). The J values are given in Hertz (Hz). Melting points were determined with Buchi apparatus and are uncorrected.
Compounds 5e7 were obtained according to the previously reported procedure and the analytical data are in accordance with the literature [16]. Compound 26 selected for behavioral evaluation was converted into the hydrochloride salt.
5.1.1.1. N-(2-nitrobenzyl)formamide (2). Nitrobenzaldehyde 1 (35 g, 1 eq) was dissolved in HCOOH (70 ml, 8 eq) and formamide (110 ml, 12 eq) was added. The mixture was stirred at 120 C for 16 h. Subsequently, the solution was diluted with CH 2 Cl 2 and washed three times with water and saturated solution of NaCl. The organic phase was dried over Na 2 SO 4 , filtered and evaporated under reduced pressure. The residue was added AcOEt to precipitate a     (3). Compound 2 (20 g, 1 eq) was dissolved in 300 ml of CH 2 Cl 2 followed by addition of TEA (77 ml, 5 eq) and POCl 3 (10 ml, 1 eq). The mixture was stirred at 0 C for 30 min. The solution was then diluted with CH 2 Cl 2 and washed with saturated solution of NaHCO 3 and NaCl. The organic phase was dried over Na 2 SO 4 , filtered and evaporated under reduced pressure. The crude product was purified by chromatography on silica gel using CH 2 Cl 2 as developing solvent. Yellow solid, yield 83%, t R ¼ 3.31 min, Mp 147e149 C,  (4). Methyl propiolate (4 ml, 1.5 eq) and Ag 2 CO 3 (0.8 g, 0.1 eq) were suspended in dioxane and compound 3 (4.8 g, 1 eq) dissolved in dioxane was added dropwise into the reaction mixture. The mixture was stirred at 80 C for 30 min. Subsequently, the solution was filtered and the filtrate was evaporated under reduced pressure. The crude product was purified by chromatography on silica gel using AcOEt/Hex 4/6 as a developing solvent.

General procedure for preparation of compounds 8ae8h
Compound 7 (0.28 mmol, 1 eq) was dissolved in CH 2 Cl 2 (5 ml), and BTPP (170 mL, 0.56 mmol, 2 eq) was added, followed by addition of respective benzyl chloride (1.8 eq). The reaction mixture was stirred for 12 h. Subsequently, the mixture was evaporated, and the remaining crude product was purified by chromatography on silica gel using AcOEt/Hex as a developing solvent.

General procedure for preparation of compounds 9e12 and 24
Compound 8 (0.3 mmol, 1 eq) was mixed together with Pd 2 (dba) 3 (0.02 eq), BINAP (0.04 eq), and NaOt-Bu (1.4 eq). The solids were suspended in 4 ml of a mixture of dioxane and tertbutanol (3/1, v/v), and respective amine (1.2 eq) was added. The reaction was irradiated with microwaves at 120 C for 1 h under argon atmosphere. The resulting mixture was filtrated through a pad of Celite, concentrated and purified by column chromatography on silica gel using CH 2 Cl 2 /MeOH as a developing solvent.

General procedure for preparation of compounds 13e23 and 25e33
Compound 8 (1.7 mmol, 1 eq) was dissolved in acetonitrile followed by addition of TEA (5.1 mmol, 3 eq) and respective amine (6.8 mmol, 4 eq). The reaction was heated in a microwave reactor at 140 C for 7 h. The solvent was evaporated and the crude product was purified by chromatography using silica gel with CH 2 Cl 2 /MeOH as a developing solvent.   13

Evaluation of water positions
Five complexes (PDB ID: 2v60, 37 2v61, 37 2v5z [37], 6fw0, 38 and 6fwc [38]) of MAO-B, crystalized with non-covalent inhibitors (which contain F or Cl atom in phenyl ring) were selected from the protein data bank (PDB) [39]. Each complex contained water molecules close to the ligand. To select the conserved water molecules in the binding site, an alignment of all complexes was performed using SiteMap with aligning residues within 5 Å from the ligand [40,41]. It was found, that four water molecules localized close to the FAD molecule, and three water molecules in the binding crevice formed by ILE199, GLN206, and TYR326 had common positions.

Induced fit docking
The 3-dimensional structures of the ligands were prepared using LigPrep v3.6 [42] and the appropriate ionization states at pH ¼ 7.0 ± 0.5 were assigned using Epik v3.4 [43,44]. Compounds with unknown absolute configuration were docked in R and S configurations. The Protein Preparation Wizard [45] was used to assign the bond orders, appropriate amino acid ionization states, and to check for steric clashes for the MAO-B crystals. The water molecules, except those that were chosen as conserved, were removed. The receptor grid was generated (OPLS3 force field) [46e48] by centering the grid box with a size of 8 Å on noncovalently bonded inhibitors. Automated flexible docking was performed using Glide v6. 9[49À51] at the SP level.

Molecular dynamics
The starting conformation of the L-R complexes were carefully selected for each ligand with a binding mode similar to the one observed in the crystal structure and which value of the RMSF of oxygen atoms in water molecules was the smallest. A 50 ns-long Molecular Dynamics (MD) simulations were performed using Schr€ odinger Desmond software. The system was solvated by water molecules described by the TIP4P potential and the OPLS3 force field parameters were used for all atoms. 0.15 M NaCl was added to mimic the ionic strength inside the cell. The output trajectories were hierarchically clustered into 5 clusters according to the ligand using trajectory analysis tools available in the Maestro Schr€ odinger Suit.

Quantum polarized ligand docking
The top-three 26-MAO-B complex clusters were selected and used as a grid for the next steps. The receptor grids were generated (OPLS3 force field) by centering the grid box with a size of 8 Å on the 26 compound. Docking was performed by quantum-polarized ligand docking (QPLD) procedure [52] involves the QM-derived ligand atomic charges in the protein environment at the 3-21G/ BLYP level. For each ligand, the 5 poses were obtained.

Binding free energy calculations
GBSA (Generalized-Born/Surface Area) was used to calculate the binding free energy based on the ligandereceptor complexes generated by the QPLD procedure. The ligand poses were minimized using the local optimization feature in Prime, the flexible residue distance was set to 6 Å from a ligand pose, and the ligand charges obtained in the QPLD stage were used. The energies of complexes were calculated with the OPLS3e force field and Generalized-Born/Surface Area continuum solvent model. To assess the influence of a given substituent on the binding, the DDG was calculated as a difference between binding free energy (DG) of the unsubstituted phenyl ring and its halogenated analog.

In vitro pharmacological evaluation
5.3.1. 5-HT 6 Rs affinity evaluation 5.3.1.1. Cell culture and preparation of cell membranes for radioligand binding assays. HEK293 cells with stable expression of human 5-HT 6 , receptors (prepared with the use of Lipofectamine 2000) were maintained at 37 C in a humidified atmosphere with 5% CO 2 and grown in Dulbecco's modified Eagle medium containing 10% dialyzed fetal bovine serum and 500 mg/ml G418 sulfate. For membrane preparation, cells were sub-cultured in 150 cm 2 flasks, grown to 90% confluence, washed twice with phosphate buffered saline (PBS) prewarmed to 37 C, and pelleted by centrifugation (200Âg) in PBS containing 0.1 mM EDTA and 1 mM dithiothreitol. Prior to membrane preparation, pellets were stored at À80 C. The process of equilibration was terminated by rapid filtration through Unifilter plates with a 96-well cell harvester, and radioactivity retained on the filters was quantified on a Microbeta plate reader (PerkinElmer, USA). For displacement studies, the assay samples contained as radioligands (PerkinElmer, USA) 2 nM [ 3 H]-LSD (83.6 Ci/mmol). Nonspecific binding was defined with 10 mM methiothepine. Each compound was tested in triplicate at 7 concentrations (10 À10 to 10 À4 M). The inhibition constants (K i ) were calculated from the ChengÀPrusoff equation [53]. Results were expressed as means of at least two independent experiments.

Monoamine oxidase assays
Inhibition activity of evaluated compounds was measured using human recombinant MAO-B and MAO-A (Sigma Aldrich M7441 and M7316) in the fluorometric method for detecting monoamine oxidase activity. The assay was carried out in a 96-well plate. 2 ml of the appropriate concentration of tested compounds in DMSO were added to wells that contained 98 ml of enzyme dilution (0.53 U/ml) in phosphate buffer (50 mM, pH 7.4). After the 30 min of preincubation at room temperature 50 ml of the solution of 800 mM10-Acetyl-3,7 dihydroxyphenoxazine (Cayman Chemical Company 10010469) and 4 U/ml horse radish peroxidase (HRP, Sigma Aldrich P6782) was added and enzymatic reaction was started by addition of 50 ml of 800 mM p-tyramine (Alfa Aesar A12220) solution. The signal was measured after 1 h (excitation at 570 nm and emission at 585 nm) using EnSpire® multimode plate reader (PerkinElmer, Inc.). Rasagiline (1 mM) or clorgyline (1 mM) were tested as reference compounds for MAO-B and MAO-A inhibitions, respectively [54,55].

Reversibility studies
To investigate the reversibility of MAO-B inhibition compound 26, rasagiline and safinamide were tested in concentration corresponding to their IC 80 values. Experiment was carried out in a 96well plate. hMAO-B was incubated with inhibitors for 30 min then low concentration of p-tyramine (10 mM in wells) and the solution of 10-acetyl-3,7-dihydroxyphenoxazine and HRP (200 mM and 1 U/ml in wells) were added to the plate. Fluorescence signal had been measured in the microplate reader for 22 min then the concentration of p-tyramine was increased to 1 mM. After the addition of p-tyramine fluorescence was measured every 5 min for 5 h in order to monitor the enzymatic reaction product formation [54e56].

Evaluation of mode of inhibition
The mode of the inhibition was tested according to the standard test procedure using different concentrations of the substrate [56e58]. Inhibitor (compound 26) was used in three concentrations corresponding to its IC 20 , IC 50 and IC 80 values. Substrate was used in six concentrations: 0.05 nM, 0.1 mM, 0.5 mM, 1 mM, 1.5 mM, 2 mM. After the experiment velocities were calculated and put on the graph (y-axis) against the concentration of the substrate (x-axis). From the Michaelis-Menten plot V max , K M and a values were calculated for different concentrations of the inhibitor. Double-reciprocal plot (Lineweaver-Burk plot) was prepared to display the data [55e57].

In vitro metabolic stability studies
Metabolic stability of tested compound was analyzed using incubation systems, composed of: tested compound (10 mM), rat liver microsomes (RLMs, microsomes from rat male liver, pooled; 0.4 mg/ml; Sigma Aldrich), NADPH-regenerating system (NADP þ , glucose-6-phosphate and glucose-6-phosphate dehydrogenase in 100 mM potassium buffer, pH 7.4; all from Sigma Aldrich) and potassium buffer, pH 7.4. The stock solution of tested compounds was prepared in methanol (the final methanol concentration in incubation mixture does not exceed 0,1%). Firstly, all samples contained incubation mixture (without NADPH-regenerating system) were pre-incubated in thermoblock at 37 C, for 10 min. Then the reaction was initiated by the addition of NADPH-regenerating system. In control samples NADPH-regenerating system was replaced with potassium buffer. Probes were incubated for 30 and 60 min at 37 C. After addition of internal standard (pentoxifylline, 10 mM) biotransformation process was stopped by addition of perchloric acid. Next, samples were centrifuge and supernatants were analyzed using UPLC supplemented with 10% FBS (Fetal Bovine Serum) and 1% pencilinstreptomycin solution at 37 C, in a humidified atmosphere with 5% CO 2 until the cells reached appropriate confluence. Cells were seeded in 96-well plates at a density of 1* 10 4 cells per well. For protection studies, astrocytes were co-treated with the cytotoxic agent (DOX) and analyzed compounds (26, SAF, SEL) for 24 h; then the ability of compounds to protect astrocytes against DOX-induced toxicity was examined using MTT assay.

MTT assay
Cell viability was estimated by using the MTT assay, which assesses the ability of mitochondrial dehydrogenases to reduce tetrazolium salts into a colored formazan compound [60]. At the end of the incubation period, 10 ml of MTT (5 mg/ml) were added to each well. After 4 h, formazan crystals were solubilized with DMSO and optical density was measured at 570 nm by using a SpectraMax iD3 Multi-Mode Microplate Reader (Molecular Devices). Each experiment was performed in triplicate and repeated three times.

Inhibition of cytochrome P450
The luminescent CYP3A4, CYP2D6 and CYP2C9 P450-Glo™ assays and protocols were provided by Promega (Madison, WI, USA). The method is based on the conversion the beetle D-luciferin derivative into D-luciferin catalyzed by the respective CYP isoform. After addition of the luciferase, the luminescence signal is produced directly proportional to the amount of obtained by CYP's D-luciferin. The enzymatic reactions were performed in white polystyrene, flat-bottom Nunc™ MicroWell™ 96-well microplates (Thermo Scientific, Waltham, MA USA). The luminescence signal was measured with a microplate reader EnSpire (PerkinElmer, Waltham, MA USA) in luminescence mode. The signal produced without the presence of the tested compound was considered as 100% of CYP activity. The IC 50 values of reference CYP inhibitors: ketoconazole (3A4), quinidine (2D6) and sulfaphenazole (2C9) were calculated according to the manufacturer's recommendations. Each compound was tested in triplicate at the final concentrations in the range from 0.01 to 25 mM. GraphPad Prism™ software (version 5.01, San Diego, CA, USA) was used to calculate IC 50 values.

Evaluation of hepatotoxicity
HepG2 (ATCC® HB-8065™) hepatoma cell line was kindly donated by the Department of Pharmacological Screening, Jagiellonian University Medical College. The cells were cultured in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal bovine serum (FBS) (Gibco, Carlsbad, CA, USA. The HepG2 cells were seeded in 96-well transparent plates at a concentration of 1 Â 10 4 cells/well and incubated for 24 h at 37 C in a 5% CO 2 atmosphere to reach 50% of confluence. Tested compounds were diluted into a fresh growth medium and added to the cells at the final concentrations 0.1 mMe100 mM. The positive controls doxorubicin and mitochondrial toxin CCCP were also added at 1 mM and 10 mM, respectively. The cells were incubated next for 72 h. To asses cells' viability, the MTS reagent (CellTiter 96® AQueous One Solution Cell Proliferation Assay, Promega, Madison, WI, USA) was added to each well. After 4 h of incubation, the absorbance was measured using a microplate reader (EnSpire, PerkinElmer, Waltham, MA USA) at 490 nm. GraphPad Prism™ software (version 5.01, San Diego, CA, USA) was used to calculate statistical significances.

Animals and ethical statement
The purpose of this study was to evaluate the pharmacokinetic profile and distribution to the brain for compound 26 in Sprague-Dawley rats (Charles River Laboratories, Germany) after a single intravenous and intragastric administration of a compound in a dose of 3 mg/kg dissolved in colliphor. The study was performed under the approval of the I Local Ethics Committee for Experiments on Animals of the Jagiellonian University in Krakow, Poland No. 333/2019. During the habituation period the groups of 5 rats were kept in a plastic cage (252 mm Â 167 cm x 140 cm) at a controlled room temperature (22 ± 2 C), humidity (55 ± 10%), full spectrum cold white light (350e400 lx), on 12 h light/dark cycles (the lights on at 7:00 a.m., and off at 19:00 p.m.), and had free access to standard laboratory pellet and tap water.

Sample preparation
A group of 70 Sprague-Dawley rats (Charles River Laboratories, Germany), 13e15 weeks of age and weighing 220e250 g were used in the study. Blood and brain samples were obtained at 7 time points (5 min, 15 min, 30 min, 60 min, 120 min, 240 min, 480 min) after intravenous or intragastric administration of compound 26. In addition, 5 rats were administered the vehicle of the test substance (colliphor) at t ¼ 0. Rats were anesthetized by i.p. injection of 50 mg/kg ketamine plus 8 mg/kg xylazine, then blood samples have been taken from the right ventricle, and the brain after decapitation. The brain was flushed three times with saline. The plasma was separated by centrifugation (3000Âg, 10 min), then plasma and brain were stored at À80 C pending analysis.
The plasma and brain sample pretreatment procedure involved acetonitrile precipitation. A 10 ml aliquot of the IS working solution (5000 ng/ml) was added to 100 ml of the collected rat plasma sample, which was then vortex-mixed for 10 s. Thereafter, 200 ml of cooled acetonitrile was added, vortexed during 20 min, and then centrifuged (3000Âg, 10 min). The supernatant (200 ml) was then transferred to insert placed in an autosampler vial, and a 20 ml volume of this was injected onto the LC column.
Brain samples were thawed before use, and whole brain were weighted and placed in a glass mortar and pestle tissue grinder, and homogenized with an appropriate amount of phosphate buffer (pH 7.4) in 1:5 ratio. Afterward, 100 ml of tissue homogenates were transferred to new Eppendorf tubes and spiked with 10 ml of the IS working solution (5000 ng/ml).
Basic parameters characterizing the pharmacokinetic profile of 26, assuming model-independent pharmacokinetics, such as volume of distribution at steady-state, half-life, clearance, mean residence time of the compound in the body, area under the concentration-time curve, maximum concentration, time to reach the maximum concentration were determined from analysis of the concentration of the compound 26 determined at different times after administration (5 min, 15 min, 30 min, 60 min, 120 min, 240 min and 480 min) in plasma and brain homogenates. Pharmacokinetic calculations were performed using Phoenix WinNonlin software (Certara, Princeton, NJ 08540 USA).
First-order elimination rate constant (l z ) was calculated by linear regression of time versus log concentration. Next, the area under the mean serum and tissue concentration versus time curve (AUC 0/t ) was estimated using the log-linear trapezoidal rule, where C n is the concentration of last sampling of each compound. AUC 0/t ¼ X n i¼1 ððC i þ C iþ1 Þ = 2Þ , ðt iþ1 À t i Þ þ C n l z (Eq. 1) The area under the first-moment curve (AUMC 0/t ) was estimated by calculation of the total area under the first-moment curve using equation (2), where t n is the time of the last sampling. AUMC 0/t ¼ X n i¼1 ððt i , C i þ t iþ1 , C iþ1 Þ = 2Þ , ðt iþ1 À t i Þ þ ðt n , C n Þ l z þ C n l 2 z (Eq. 2) Mean residence time (MRT) was calculated as: MRT ¼ AUMC 0/t AUC 0/t (Eq. 3) Total clearance (Cl T ) was calculated as: The volume of distribution at steady-state (V ss ) was calculated as: The absolute bioavailability after intragastric administration was calculated as: where D i:v: and D i:g: . are i.v. and i.g. doses of studied compounds, respectively.
5.9. In vivo pharmacological evaluation 5.9.1. Animals NOR study: Male Sprague-Dawley rats (Charles River, Germany) weighing 250e280 g on arrival were housed in a temperature-controlled (21 ± 1 C) and humidity-controlled (40e50%) colony room under a 12/12 h light/dark cycle (lights on at 06:00 h). The rats were group-housed (5/cage) with free access to food and water. Rats could acclimatize for at least 7 days before the start of the experimental procedures. Behavioral testing was performed during the light phase of the light/dark cycle. The experiments were conducted in accordance with the European Guidelines for animal welfare (2010/63/EU) and were approved by the II Local Ethics Committee for Animal Experiments at the Maj Institute of Pharmacology, Polish Academy of Sciences, Krakow, Poland.
FST study: The experiments were performed on male Swiss albino mice (22e26 g) purchased from a licensed breeder Staniszewska (Ilkowice, Poland). The animals were kept in groups of ten to Makrolon type 3 cages (dimensions 26.5 Â 15 Â 42 cm) in environmentally controlled rooms (ambient temperature 22 ± 2 C; relative humidity 50e60%; 12:12 light:dark cycle, lights on at 8:00). They were allowed to acclimatize with the environment for one week before commencement of the experiments. Standard laboratory food (Ssniff M-Z) and filtered water were freely available. All the experimental procedures were approved by the I Local Ethics Commission at the Jagiellonian University in Krak ow. The experiment was conducted in the light phase between 09.00 and 14.00 h. Each experimental group consisted of: 7e9 animals/dose. The animals were used only once. The experiment was performed by an observer unaware of the treatment administered.

Drugs
Compound 26 was suspended in the mixture of EtOH/colliphor/ water (0.05/0.10/0.85; v/v/v) and was administered intraperitoneally (i.p.) at a volume of 10 ml/kg, 60 min before the test. Control animals received a vehicle injection.

Novel object recognition test
The test was carried out as described earlier [61]. Briefly, rats were tested in a dimly lit apparatus made of a dull gray plastic (66 Â 56 Â 30 cm). Following habituation on day 1, two trials separated by an inter-trial interval (ITI) of 1 h were carried out on the day 2. During the first trial (familiarization, T1) two identical objects (A1 and A2) were presented in the opposite corners of the apparatus, approximately 10 cm from the walls. During the second trial (recognition, T2) one of the A objects were replaced by a novel one, so that the animals were presented with the A ¼ familiar and B ¼ novel objects. Both trials lasted for 3 min. Rats were considered exploring objects when looking, licking, sniffing, or touching the object. Exploration time of the objects was measured manually by an experimenter unaware of treatment conditions. Based on exploration time (E) of two objects during T2, discrimination index (DI) was calculated according to the formula: DI ¼ (EBeEA)/ (EA þ AB).

Forced swim test
The experiment was carried out according to the method of Porsolt et al. [62] Briefly, Swiss albino mice were individually placed in a glass cylinder (25 cm high; 10 cm in diameter) containing 10 cm of water maintained at 23e25 C, and were left there for 6 min. The total duration of immobility was recorded during the last 4 min of a 6-min test session. A mouse was regarded as immobile when it remained floating on the water, making only small movements to keep its head above it.

Statistics
The data are presented as the mean ± SEM. The statistical significance of the results was evaluated by a one-way ANOVA, followed by Newman-Keuls (NOR study) and Bonferroni's Comparison Test (FST study); p < 0.05 was considered significant.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.