Through scaffold modification to 3,5-diaryl-4,5-dihydroisoxazoles: new potent and selective inhibitors of monoamine oxidase B

Abstract 3,5-Diaryl-4,5-dihydroisoxazoles were synthesized and evaluated as monoamine oxidase (MAO) enzyme inhibitors and iron chelators. All compounds exhibited selective inhibitory activity towards the B isoform of MAO in the nanomolar concentration range. The best performing compound was preliminarily evaluated for its ability to bind iron II and III cations, indicating that neither iron II nor iron III is coordinated. The best compounds racemic mixtures were separated and single enantiomers inhibitory activity evaluated. Furthermore, none of the synthesised compounds exhibited activity towards MAO A. Overall, these data support our hypothesis that 3,5-diaryl-4,5-dihydroisoxazoles are promising scaffolds for the design of neuroprotective agents.


Introduction
Amine oxidase is a widely diffused family of enzymes responsible for the oxidative deamination of both endogenous and exogenous monoamines, diamines and polyamines. These enzymes are divided into two classes, flavin adenine dinucleotide amino oxidase (FAD-AO) (EC 1.4.3.4) 1 and copper topaquinone amino oxidase (Cu/TPQ-AO) 2 . The former, named monoamine oxidase (MAO), contains flavin adenin dinucleotide (FAD) as a cofactor while the latter contains 2,4,5-trihydroxyphenylalanine quinone as cofactor (TPQ-Cu) and are inhibited by semicarbazide. Regarding FAD-AO, two isoforms, MAO-A and MAO-B, have been isolated, differing from central nervous system (CNS) location, substrate specificity and sensitivity to inhibitors 3,4 . MAO enzymes 5 play a key role in the metabolism of monoamine neurotransmitters, and have been therefore studied in a number of psychiatric and neurological diseases 6 . Moreover, the MAO catalysed oxidative deamination of amines leads to the formation of hydrogen peroxide which is co-responsible for neurodegeneration 7 . Not surprisingly a number of MAO inhibitors have been investigated as neuroprotective agents in the treatment of Parkinson and Alzheimer diseases 3,4,6,[8][9][10] . We have previously identified highly potent and selective MAO B inhibitors characterised by a 3,5-diaryl-4,5-dihydro-(1H)-pyrazole scaffold 11 and investigated on the structural requirements for the enzyme inhibition ( Figure 1). In particular, the presence of two aryl substituents on the dihydro-(1H)-pyrazole nucleus appeared as essential for MAO B selectivity and inhibition potency activity [12][13][14][15] . Hence, we designed and studied 2-thiazolylhydrazones derivatives, ideally originated by removing C4 and C5 from the dihydropyrazole ring. These derivatives exhibited high potency and selectivity towards MAO B [16][17][18][19] . More recently, through the isosteric replacement of the carbon 4 of the previously reported dihydropyrazole derivatives 11 with an oxygen atom, we have designed and synthesised two series of 2,3dihydro-oxadiazoles 20-22 that exhibited high selectivity and activity towards MAO B. Moreover, we observed that the introduction of a 3,4-dichlorophenyl substituent in the position 2 of the dihydro-oxadiazole core is crucial for both selectivity and potency 21 . According to these structural information and with the aim of setting new hit structures for MAO B inhibition, we have devised a synthetic pathway to obtain 3,5-diaryl-4,5-dihydroisoxazoles as bio-isosters of both the previously reported 3,5-diaryl-4,5-dihydro-(1H)-pyrazoles and 2,5-diaryl-2,3-dihydro-1,3,4-oxadiazoles.

Materials and apparatus
Starting materials and reagents were obtained from commercial suppliers and were used without purification. All melting points were determined on a Stuart SMP11 melting points apparatus and are uncorrected (East Yorkshire, UK). Electron ionization mass spectra were obtained by a Fisons QMD 1000 mass spectrometer (70 eV, 200 mA, ion source temperature 200 C) (Loughborough, UK). Samples were directly introduced into the ion source. Found mass values are in agreement with theoretical ones. Melting points, yield of reactions and the analytical and descriptive data of derivatives are reported in Table 1. 1 H-NMR (Table 2) were registered on a Bruker AMX 300 MHz spectrometer (Billerica, MA). All samples were measured in CDCl 3 . Chemical shifts are reported referenced to the solvent in which they were measured. Coupling constants J are expressed in Hz. Elemental analyses were obtained on a Perkin-Elmer 240 B microanalyser (Waltham, MA). Analytical data of the synthesised compounds are in agreement within ±0.4% of the theoretical values.
TLC chromatography was performed using silica gel plates (Merck F 254, Kenilworth, NJ), spots were visualised by UV light.

General procedure for the synthesis of compound derivatives
In a typical experiment, 0.0038 mol of the appropriate a,b-unsaturated ketone is dissolved in 30 ml of boiling absolute ethanol. On this solution, 10 ml of a freshly prepared hydroxylamine hydrochloride (0.0057 mol) and potassium hydroxide (0.013 mol) water solution are added dropwise. The solution becomes progressively orange and the reaction is monitored by thin layer chromatography (TLC) (CH 2 Cl 2 /n-hexane: 20/1). By cooling the reaction, the formation of a precipitate is observed which was filtered and crystallised from ethanol.

HPLC enantioseparations and stereochemical characterisation
Direct separation of the enantiomers of EMAC II (i-m) was achieved by HPLC using pure dichloromethane as the eluent. The resolving power of the immobilized-type was sufficiently high to achieve a baseline enantioseparation of all compounds in a shorttime analysis. The optimized analytical conditions were easily scaled up to semipreparative level employing a 1 cm i.d. IA column. An amount of <20 mg of racemic samples was resolved for each chromatographic run, and both enantiomers were collected with high enantiomeric purity. The stereochemical characterization of compounds EMAC II (i-m) was performed by CD correlation comparing the maximum and minimum of ellitipticity of the CD spectra of the isolated enantiomers of compound EMAC II i of known stereochemistry (Figure 2), as previously described by some of us 23 . Single enantiomers were resubmitted to biological assay.

Biological assay
The  24 Hence, reported IC 50 is a useful tool to determine the relative activity of the compounds within the series as well as to determine substituent and key positions in the scaffold. All the tested compounds exhibited inhibitory activity towards hMAO-B at nM concentration. Single enantiomers activity was evaluated indicating that stereochemistry plays a relevant role in determining inhibitory concentrations. None of the tested compounds exhibited activity towards the A isoform of the enzyme at the tested concentration of 100 mM.

Coordination studies
The interaction between EMAC II l and the Fe 2þ or Fe 3þ ions was evaluated by recording the UV/VIS absorbance spectra of EMAC II l (Figure 3) in absence and in presence either of FeCl 3 or FeSO 4 . EMAC II l was solubilised in DMSO then diluted to 1 mM in

Molecular modelling studies
Protein preparation. Three-dimensional coordinates of the receptor 25 were obtained from the Protein Data Bank (PDB). 26 The protein was processed and the internal hydrogen bonding network of the receptor was optimized using the algorithm implemented in Protein Preparation wizard. 27 Ligand preparation. The ligands were built using Maestro GUI 27 . The lowest energy conformer was considered for the subsequent computational studies. This was obtained with MacroModel version 7.2 28 , considering MMFFs 29 as force field and solvent effects by adopting the implicit solvation model Generalized Born/Surface Area (GB/SA) water 30  Docking experiments. Previous reported and validated QMPL docking protocol was applied using default settings 21,31 . In order to better take into account the induced fit phenomena, the most energy favoured generated complexes were fully optimized with the AMBER Ã united atoms force field 32 in GB/SA implicit water, 30 setting 5000 steps iterations analysis with PRCG method and a convergence criterion of 0.1 kcal/(mol Å). Analysis of the results of the complex minimization was carried out taking into account the state equations (free energy of complex formation) computed at 300 K, applying molecular mechanics and continuum solvation models with the molecular mechanics generalized Born 33 . The resulting complexes were considered for the binding modes graphical analysis with Pymol 34 and Maestro 27 .

Results and discussion
Pursuing in our research on the design and synthesis of new agents for the selective inhibition of MAO B, we have synthesised a series of 3,5-diaryl-4,5-dihydro-isoxazoles EMAC II (i-m). All the  compounds were submitted to biological evaluation to investigate their ability to inhibit both MAO isoforms. The structure and the MAO A and MAO B IC 50 values and selectivity ratios of the new derivatives are reported in Table 3. All the compounds inhibit MAO B isozyme with IC 50 values ranging from higher to lower nanomolar range. Accordingly to what is observed in our previous report on similar derivatives 21 , compound EMAC II l, bearing a 3,4-dichlorophenyl moiety in the position 5 of the dihydroisoxazole ring resulted, the most active within all the tested compounds. Based on this observation we assumed that, in the formation of the enzyme-ligand complex, this portion of the inhibitor is, in all probability, oriented similarly with respect to our previously reported compounds. Interestingly, none of the new compounds exhibited activity towards the MAO A isoform of the enzyme up to 100 mM concentration. Considering the presence of an asymmetric carbon at the position 5 of the dihydro-isoxazole nucleus we performed, semipreparative chromatographic enantioseparation of all the new derivatives EMAC II (i-m). The separated enantiomers were then submitted to biological evaluation ( Table  4). The obtained data pointed out the importance of enantiomeric separation for both potency and selectivity. In particular, in the case of compound R-(À)-EMAC II l, the IC 50 value towards MAO B, drops from 11.97 nM to 1.89 nM, while the selectivity towards MAO B is increased by more than six folds. Moreover, considering the growing interest in the development of iron chelators for the treatment of neurodegenerative disorders [35][36][37] , we have preliminarily investigated our best compound able to bind iron II and iron III cations following a previously reported procedure 38 . According to our preliminary results (Figure 3), compound EMAC II l do not have the ability to bind neither iron II nor iron III cations.
In order to explore the nature of the ligand-receptor interactions, we have carried out docking experiments considering the MAO B active site by means of QMPL docking protocol 31,39,40 . We have focused our attention on the most active compound R-(À)-EMAC II l, which shows the lowest IC 50 value and the highest selectivity towards MAO-B.
It should be pointed out that MAO-B cavity is rather narrow, therefore, small changes are reflected in the biological activity. In this respect, the best compound shows remarkable difference of activity between the R and the S enantiomers. An analogous behaviour was observed in the previous series 11,20 . According to docking and post docking best scored pose, the most stable complex configurations of both R and S enantiomers are depicted in Figure 4(a,b).  The R enantiomer is stabilized by pÀp interactions with Trp119, Phe103, Phe343, Tyr326 and some hydrophobic interactions with several residues in the binding pocket: Leu 164, Leu 167, Phe 168, Ile 316, Ile 199, Ile198, Leu 171 and the Tyr of the aromatic cage 435, 298, 188.
Conversely, the S enantiomer is oriented in the opposite way: with the dichlorophenyl substituent in the entrance cavity and the 4-methylphenyl substituent in the catalytic cavity toward the FAD. The complex results in less stability and gives rise only to a single pÀp interaction with Tyr 435 and several hydrophobic interactions (Figure 4(c,d)).
Compared with previous series 21 , the absence of acetyl group allows a better accommodation in the cavity (Figures 4(a) and 5(a)). Once again, the importance in crossing both entrance and catalytic cavity, in order to achieve the MAO-B isoform selective activity is confirmed. This essential requisite has been reported by several studies 25,41,42 and can be observed in co-crystallized selective MAO B inhibitors. A typical example is the inhibitor safinamide ( Figure 5(c,d)) 25,43 .

Conclusions
In summary, a small library of 3,5-diaryl-4,5-dihydroisoxazoles have been designed and synthesized on the basis of previously reported MAO inhibitors. Our finding highlighted the capability of 3,5-diaryl-4,5-dihydroisoxazoles to exert specific recognition and inhibition against the isoform B of MAO. The stereochemical characterization of R/S enantiomers confirms an enantioselective effect against MAO B isoform. This scaffold is definitively interesting for the development of novel selective neuroprotective agents.

Disclosure statement
The authors have no declarations of interest to report.