Molecular Modeling and Anticholinesterasic Activity of Novel 2-Arylaminocyclohexyl N , N-Dimethylcarbamates

O presente trabalho reporta um estudo teórico e experimental detalhado das séries inéditas de isômeros cise trans-N,N-dimetilcarbamatos de cicloexila 2-arilaminossubstituídos como potenciais inibidores de colinesterases. Os testes de inibição in vitro, realizados através do método de Ellman em amostras de sangue humano, mostraram que os novos carbamatos apresentaram boa seletividade frente à inibição da enzima butirilcolinesterase (BuChE), com um máximo de inibição de 90% e IC50 de 6 e 8 mmol L -1 para os compostos mais ativos da série. Os estudos de modelagem molecular apontaram significantes diferenças entre as orientações destes compostos nos sítios ativos das enzimas BuChE e acetilcolinesterase (AChE). Os resultados mostraram que os compostos interagem de forma mais efetiva com o sítio ativo da enzima BuChE, pois o grupo carbamato está próximo aos resíduos chave da tríade catalítica.


Introduction
Alzheimer's disease (AD) is a neurodegenerative disease characterized by progressive loss of memory and other functions.It was first described by Alois Alzheimer in 1906. 1 This type of mental deterioration is the most common in individuals aged over 60 and its incidence has increased as the world population has grown older.Currently, it is estimated that 26 million people suffer from AD in the world. 2 Despite the numerous studies done, the causes of AD are not fully elucidated and no cure has been found yet.Nevertheless, its damages and progression can be minimized.][5][6] In physiological conditions, the activity of neurotransmitter ACh is known to be related mainly to acetylcholinesterase (AChE).[9][10] Recent studies have shown that due to a progressive decrease in the AChE activity in certain regions of the brain in AD, 4,11 the BuChE activity increases, suggesting that BuChE may act as a compensatory mechanism in the hydrolysis of ACh. 12,131][22][23][24] Rivastigmine, a drug from the carbamate class, is one of four drugs that act as cholinesterase inhibitors currently approved by ANVISA, the Brazilian Health Surveillance Agency, for the treatment of AD.
This study describes the synthesis and evaluation of the in vitro activity of two novel series of potential cholinesterase inhibitors (Figure 1).Computational calculations were performed to determine the conformational preference of the isomers and molecular docking was performed for the proposal of a binding model for the compounds synthesized to the active site of enzymes AChE and BuChE.We hope this study may contribute to increase the number of therapeutic agents against AD.

Synthesis
New cis-and trans-carbamate derivatives were synthesized via the route outlined in Scheme 1. trans Carbamates were synthesized in two steps.First, cyclohexene oxide was submitted to aminolysis with arylamines 25 to obtain trans-2-arylaminocyclohexanols 1a-1c with yields from 50 to 80%.Next, the alcohols 1a-1c were carbamoylated 26 with dimethylcarbamoyl chloride to give carbamates 2a-2c with yields ranging from 35 to 50%.
The cyclohexene oxide aminolysis was performed without a catalyst and with water as a solvent.Bonollo et al. 25 demonstrated that the reaction efficiency depends on a rigorous pH control.The ideal reactional condition is attained when 1.0 mmol cyclohexene oxide is mixed with 1.05 mmol amine in 2.0 mL deionized water; the resulting pH of 8.30 varies little during the reaction.
The trans-2-arylaminocyclohexanols 1a-1c were carbamoylated with metallic sodium for the preparation of the corresponding alkoxide and later reacted with dimethylcarbamoyl chloride.However, the joint analysis of the nuclear magnetic resonance (NMR) and gas chromatography-mass spectrometry (GC-MS) data in the preparation of derivative 2c showed the formation of a mixture of compounds, 25% of the defluorinated compound 2a and 75% of the desired product 2c.Attempts to separate these compounds by column chromatography were unsuccessful.As an alternative, metallic sodium was substituted with sodium hydride in the synthesis of compound 2c, which was obtained as a single product.The mechanism for hydrodefluorination was reported previously 27,28 as a sequence of steps, beginning with the electron transfer from metal to the substrate and ending with the formation of the hydrodefluorination product.Therefore, the product 2c was successfully obtained when the metallic sodium was substituted by sodium hydride since in this case it is not possible the occurrence of electron transfer.
The cis carbamates were synthesized in four steps.First, cyclohexanone was reacted with bromide to produce 2-bromocyclohexanone 29 with a yield of 70%.Next, racemic 2-arylaminocyclohexanones 3a-3c were prepared with 30% yield by substitution of 2-bromocyclohexanone with appropriate arylamines, 30 using quinoline as a catalyst and ethoxyethanol as a solvent.In the third step, ketones 3a-3c were stereoselectively reduced in the presence of N-selectride, 31 giving the respective cis-2-arylaminocyclohexanols 4a-4c in moderate yields, 36-50%.Temperature control (-78 o C) was very important in this step to ensure that only the cis isomer was obtained.Finally, compounds 4a-4c were carbamoylated 26 under the same conditions used in the synthesis of the trans carbamates to obtain compounds 5a-5c.

Biological activity
Inhibitory potencies of the novel synthesized carbamates against cholinesterases from fresh human blood were evaluated by Ellman's modified spectroscopic method. 32his test is based on the reaction of 5,5'-dithiobis-(2-nitrobenzoic acid), known as DTNB or as Ellman's reagent, with the sulfhydryl group of acetylthiocholine or butyrylthiocholine, which results in the formation of a yellow-colored product, 2-nitro-5-thiobenzoic acid (TNB).This product has maximum absorbance at 412 nm, thus the lower the measured absorbance, the greater the activity of the compounds tested, since the enzyme is inhibited.Since this method requires water soluble compounds, chlorohydrates 2a'-2c' and 5a'-5c' were prepared.
The assays were performed at five concentrations (0.01, 0.025, 0.05, 0.1 and 0.2 mol L -1 ) of the potential inhibitors.Exelon ® , a drug that contains the active principle rivastigmine, was used as a standard.In these experiments, the inhibition potential of each compound against butyrylcholinesterase (found in plasma) and acetylcholinesterase (found in red blood cells) was assessed.
The inhibition results are shown in Figure 2a-2d.The IC 50 (the testing compound concentration that inhibits the hydrolysis of substrates by 50%) values (Table 1) of each compound were obtained through the plot of percent inhibition versus compound concentration.
The analysis of Figure 2 shows that all compounds were more active in the butirylcholinesterase inhibition for any evaluated concentration (Figures 2a and 2c).The cis and trans salts produced maximum activity inhibition for BuChE, around 90% at 0.2 mol L -1 .Compounds 5a' and 2b' were the most active, with IC 50 values of 6.0 and 8.0 mmol L -1 , respectively.
Relative to the acetylcholinesterase inhibition (Figures 2b and 2d), compounds 2a'-2c' and 5b'-5c' followed the same tendency and did not inhibit 50% of the enzyme activity in any of the investigated concentrations.Only compound 5a' inhibited AChE significantly, with a maximum inhibition of 79% at 0.2 mol L -1 .
Likewise, the investigated compounds had good selectivity in relation to the BuChE inhibition.Derivative 2b' was the most active selective compound, with an IC 50 of 8.0 mmol L -1 .Among the tested compounds, only derivative 5a' was active against both enzymes, with an IC 50 of 6.0 mmol L -1 against plasma cholinesterase (BuChE) and 59.2 mmol L -1 against erythrocyte cholinesterase (AChE).
The analysis of the data shows that the cis-trans isomerism does not influence the activity of the compounds significantly.However, the presence of donor or acceptor substituents in the para position of the aromatic ring may either increase or decrease the activity of the compounds, depending on the series that is analyzed.For the trans series (compounds 2a'-2c'), the presence of the methoxyl group in 2b' potentialized the anticholinesterasic activity and was approximately two times more active than derivative 2a'.For the cis series, the presence of activator (5b') and deactivator (5c') substituents in the aromatic ring significantly reduced the activity of the compounds in relation to the non-substituted derivative 5a'.
Bocca et al. 23 studied the inhibitory properties of 2-N,N-dimethylaminecyclohexyl 1-N',N'-dimethyl-carbamate isomers and their methyl sulfate salts, and found an IC 50 of 59 mmol L -1 for the most active compound of the series against BuChE.Comparatively, all compounds tested in this work (2a'-2c' and 5a'-5c') were more active against BuChE, with IC 50 values ranging from 6 to 29.8 mmol L -1 .These data demonstrate that the presence of the arylamine group potentialized the activity of these carbamates in the inhibition of plasma cholinesterase.

Conformational equilibrium analysis
The molecular modeling calculations were performed with Gaussian 03 package programs. 33To investigate the cis and trans isomer preferential conformations and the influence of the substituents on the conformational equilibrium, surface energy potential calculations were performed at HF/6-31G level.The most stable structures were then optimized at B3LYP/6-311++G(d,p) level, and the obtained energy values were used to determine the contribution of each conformer to the equilibrium.Table 2 shows the relative energies and calculated populations of each conformer of the studied compounds.Figure 3 presents the optimized structure for the cis and trans carbamate derivates of aniline (2a and 5a).Similar conformations were obtained for the other compounds.
According to Table 2, all trans isomers presented ee as major conformers.This preference can be attributed to   steric repulsion between the individual substituents and the hydrogens in 1 and 3 arrangements (syn, 1,3-diaxial interactions).When the substituents adopted equatorial positions (ee), these repulsions were avoided and otherwise, when the substituents adopted axial positions (aa), these interactions were maximized.These results agree with the coupling constant values 3 J HH obtained from the 1 H NMR spectra, with 3 J HH greater than 9.3 Hz for hydrogens H-1 and H-2, indicating that they are in axial position and the substituents (carbamate and arylamine) in equatorial position.
For isomers cis-2-arylaminocyclohexyl N,N-dimethylcarbamates, two conformations are possible, one with the carbamate group in axial position and the arylamine group in equatorial position (ae), and another with the carbamate group in equatorial position and the arylamine group in axial position (ea).Each of the conformations had two minimum energy structures, represented by ae1, ae2, ea1 and ea2 (Figure 3).The analysis of the data in Table 2 shows that the conformers with the carbamate group in the axial position (ae) have lower energy than the conformers with the arylamine group in axial position (ea).These results agree with the values of the coupling constant 3 J HH obtained from the 1 H NMR spectra.In this case, values of 3 J HH over 9.3 Hz were observed only for hydrogen H2, indicating that it is in axial position and, consequently, the arylamine group is in equatorial position.
The preference of conformer ae can also be explained by the steric repulsion between the individual substituents and the hydrogens in 1 and 3 arrangements.In the ae conformation, the atoms of the carbamate group (despite its volume being greater than that of the arylamine group) are relatively distant from the cyclohexane ring, which decreases the syn 1,3-diaxial interactions.In contrast, in the ea conformation, the arylamine group is close to the ring, causing a greater repulsion between the hydrogen in the amine group and the 1,3-diaxial hydrogens.These results are consistent with previous studies published on these compounds. 34,35he effect of hyperconjugative interactions on conformers ae and ea was assessed by means of NBO analysis, using B3LYP/6-311++G(d,p) theory level.
Table 3 shows the main interactions involving the substituent groups.When the carbamate group is in axial position (ae), a stabilizing orbital interaction occurs between σ C2-H2' → σ* C1-O , which has considerably more energy than the interaction between σ C1-H1' → σ* C2-N , which occurs when the arylamine group is in axial position (ea).This occurs because antibonding orbital σ* C1-O is a better electron acceptor than antibonding orbital σ* C2-N .The sum of the main hyperconjugative interactions of the substituent groups shows that they are more effective in the ae conformation, with a difference of about 1.50 kcal mol -1 in relation to the ea conformation.The conformational preference of the cis isomers can be assigned to both steric and electronic effects.
The molecular modeling calculations were also applied to the protonated form of the carbamates.The differences of energy between the conformers (E ee -E ea and E ae -E ea ) were higher than 5.0 kcal mol -1 , so ee and ae are the dominant conformers in the equilibrium.

Enzyme-inhibitor interactions
Molecular docking calculations were performed using the AutoDock 4.2.3 program implemented at the interface PyRx 0.9 36 to assess the enzyme-inhibitor interactions and propose a bonding model based on the experimental results.All the experimentally tested compounds were docked to both enzymes (AChE and BuChE) and their interactions were analyzed.Due to the similarity of the obtained results, the most active selective compound 2b' was chosen as model in the present discussion.It is important to point out that only the most stable conformer of each isomer was used in the docking calculations.
Figure 4a shows the trans-2-(4-methoxyphenylamino) cyclohexyl N,N-dimethylcarbamate hydrochloride 2b' complexed with enzyme AChE.One can observe that the ligand is stabilized by residues of the catalytic anionic site Glu199 and Phe330 and by one of the residues of the anionic subsite, Gly118.There is a hydrogen bond between Hsd440 and the protonated amino group.Residues Trp84, Tyr334 and Tyr121 contribute to the stabilization through van der Waals interactions.The hydrolysis mechanism of acetylcholine involves mainly the catalytic triad residues (Ser200, His440 and Glu327).Upon analysis of the enzyme-substrate complex, one can observe that the carbamate group is distant from the catalytic triad residues, decreasing the possibility of the enzyme inhibition.
Figure 4b shows the ligand complexed with enzyme BuChE.In this case, one can observe that the carbamate group is close to the catalytic triad residues of the enzyme (Ser198, Hsd438 and Glu325), enabling the nucleophilic attack of the oxygen of Ser198 to the carbonyl of the carbamate group.It can also be observed that the arylamine group is stabilized by hydrogen bonds with Glu197, Gly116 and Ser198, and those residues Tyr128, Trp82 and Tyr440 contribute to stabilize the ligand through van der Waals interactions.
In general, the molecular docking results show that the assessed compounds are located in the active site of both enzymes, but in butyrylcholinesterase, the carbamate group is close to the catalytic triad residues, whereas in acetylcholinesterase this group is farther away.These results agree with the experimental data, which showed that the tested compounds are selective inhibitors of enzyme butyrylcholinesterase.

Conclusions
New selective inhibitors of enzyme BuChE may serve as biological investigative tools of its role in AD and its treatment.This study resulted in a series of new cholinesterase inhibitors, selective BuChE inhibitors (2a'-2c', 5b' and 5c'), and one non-selective AChE/BuChE inhibitor (5a').The in vitro inhibition test by Ellman's modified method showed that the cis-trans isomerism does not influence the activity of the compounds significantly, but that the presence of donor and acceptor substituents in the para position of the aromatic ring may either increase or decrease the activity of the compounds, depending on the series being analyzed.
According to the DFT (density functional theory) calculations, the trans carbamates presented diequatorial substitution as major conformers, which can be attributed to steric repulsion between the substituents and the hydrogens in 1 and 3 arrangements (syn 1,3-diaxial interactions).For the cis isomers, the most stable conformer is the one with the carbamate group in axial position and the arylamine group in equatorial position.In this case, the conformational preference is explained both by steric and electronic effects.Molecular modeling studies attributed the selectivity of the compounds toward BuChE to the distance between the carbamoyl group and the catalytic site of the enzyme.In the BuChE, the carbamoyl group is oriented towards the catalytic triad, creating import interactions between the triad and the active site.In contrast, in AChE, the long distance between the carbamoyl group and the catalytic triad of enzyme makes difficult any interaction between them.Overall, these results indicate that the novel carbamate derivatives are interesting structures for the development of selective and more potent BuChE inhibitors.

Experimental
General experimental procedures NMR spectra were acquired in a Varian Mercury Plus BB apparatus operating at 300.059 MHz for 1 H and 75.457MHz for 13 C.The spectra were recorded in 20 mg cm -3 solutions of CDCl 3 , with a probe temperature of ca.300 K and TMS (tetramethylsilane) as reference.Melting points were determined with a Micro-Química apparatus model MQAPF-301 and are uncorrected.
HRMS analyses were performed for new 2-arylaminocyclohexyl N,N-dimethylcarbamates.The compounds were dissolved in a solution of 50% (v/v) chromatographic grade acetonitrile (Tedia, Fairfield, OH, USA), 50% (v/v) deionized water and 0.1% formic acid.The solutions were infused directly and individually into the ESI (electrospray ionization) source by means of a syringe pump (Harvard Apparatus) at a flow rate of 10 μL min −1 .ESI(+)-MS and tandem ESI(+)-MS/MS were acquired using a hybrid highresolution and high accuracy (5 μL L -1 ) MicrOTOF-Q II mass spectrometer (Bruker Daltonics) under the following conditions: capillary and cone voltages were set to +3500 and +40 V, respectively, with a de-solvation temperature of 200 o C. For data acquisition and processing, Micro-TOF software (Bruker Daltonics) was used.The data were collected in the m/z range of 50-400 at the speed of two scans per s, providing the resolution of 50,000 (FWHM) at m/z 200.
Procedures for the synthesis and purification of all compounds are described in the Supplementary Information section.

Cholinesterase inhibition bioassay
The inhibitory activity of cholinesterases (AChE and BuChE) was evaluated in fresh human blood by Ellman's modified spectrophotometric method. 32For measurement, 10 mL of a 100 mmol L -1 phosphate buffer solution pH 8.0 and 10 μL of heparinized fresh blood were used.
To determine the total amount of cholinesterases, 3.0 mL of solution (erythrocyte + plasma) were transferred to vials and 40 μL of tested compounds were preincubated with the enzymes for 10 min at 30 ºC before starting the reaction by adding of the substrate (acetylthiocholine iodide).Next, 50 μL of 10 mmol L -1 DTNB (Ellman's reagent, Sigma-Aldrich Co.) solution and 20 μL of 75 mmol L -1 acetylthiocholine iodide (Sigma-Aldrich Co.) solution were added.Enzyme activity was determined by measuring the absorbance at 412 nm for 5 min with a Shimadzu UV-1061PC apparatus.An inhibitor-free sample was used (100% enzyme activity) as a reference.Each compound was assayed at five concentrations in triplicate.The reaction rates were compared and the percent inhibition due to the presence of the tested compounds was calculated.The IC 50 values were determined by spectrophotometric measurement of the effect of increasing compound concentrations on the enzyme activity.
To determine the plasma cholinesterase (BuChE), heparinized fresh blood was centrifuged at 2000 rpm for 10 min for the sedimentation of red blood cells, and 20 μL of the supernatant (plasma) were added to 12 mL of a 100 mmol L -1 phosphate buffer solution at pH 8.0.Amounts of 3.0 mL of this solution were transferred to vials and 40 μL of test compounds were preincubated with the enzymes for 10 min at 30 ºC.Next, 25 μL of 10 mmol L -1 DTNB solution and 20 μL of 75 mmol L -1 acetylthiocholine iodide solution were added.Enzyme activity was measured as for total cholinesterase.

Ligand modeling
The molecular modeling calculations were performed with Gaussian 03 package programs. 33Energy potential surfaces were obtained for all the compounds at theory level HF/6-31G to determine the position of lowest energy adopted by the groups in space.The lowest energy structures were optimized at a high level of theory [B3LYP/6-311++G(d,p)] and characterized as absolute minima by vibrational frequency calculations.NBO calculations (version 5.0) were performed to evaluate the hyperconjugative interactions using theory level B3LYP/6-311++G(d,p).

Molecular docking
Molecular docking studies were performed using the AutoDock 4.2.3 program implemented at the interface PyRx 0.9. 36The compounds under study were docked to 1GQR crystallographic structures, AChE complexed with rivastigmine, and 1POM, BuChE complexed with choline ions, selected from the Protein Data Bank (PDB).For each PDB file, molecules of water and other ligands (except the main ligands rivastigmine and choline ion) were removed.Redocking calculations were performed to validate the parameters that had been chosen.The compound structures were drawn and optimized with the Gaussian 09 program 33 as described in the Ligand Modeling section.
The docking calculations for the enzymes and compounds were performed by building a 50 × 50 × 50 Å box centered at coordinates x = 8.058, y = 64.887and z = 61.330for AChE and x = 132.478,y = 114.974and z = 38.835for BuChE with a grid space of 0.375 Å.The genetic algorithm (GA) was used as a standard protocol of 50 poses obtained for the ligand, an initial population of 150 random individuals, a maximum number of 2.5 × 10 5 energy evaluations and a maximum of 2.7 × 10 4 generations.The docked results within an RMSD (root mean square deviation) of 2.0 Å were clustered and the final results of each ligand were selected considering both the embedded empirical binding free energy evaluation and the clustering analysis.
The best results were submitted to energy minimization with the NAMD2 program. 37The force field adopted for proteins was CHARMM C35b2-C36a2, and for the ligands, they were generated in the same format by the SwissParam server. 38Energy minimization was simulated with the complexes immersed in a box with water measuring at least 10 Å from the outermost surface of the protein.Either Na + or Cl -counter ion was added in appropriate amounts to neutralize the charges of the system.The temperature and the pressure were adjusted to 300 K and 1 atm.After energy minimization, the protein-ligand complexes were redocked using the same docking parameters, which gave a maximum RMSD of 1.5 Å.

Figure 1 .
Figure 1.Structures of the two novel series of carbamate derivatives cis-and trans-2-arylaminocyclohexyl N,N-dimethylcarbamates.

Table 1 .
IC 50 values of carbamate chlorohydrates and reference compound, rivastigmine, as inhibitors of cholinesterases