Design, synthesis and biological evaluation of novel antipyrine based α-aminophosphonates as anti-Alzheimer and anti-inflammatory agent

Abstract Herein, a series of novel antipyrine based α-aminophosphonates derivatives were synthesized and characterized. The synthesized derivatives were subjected for in vitro cholinesterase inhibition, enzyme kinetic studies, protein denaturation assay, proteinase inhibitory assay and cell viability assay. For cholinesterase inhibition, the results inferred that the test compounds possess better AChE activity (0.46 to 6.67 µM) than BuChE (2.395 to 12.47 µM). Compound 4j inhibited both AChE and BuChE (IC50 = 0.475 ± 0.12 µM and 2.95 ± 0.16 µM, respectively), implying that it serves as a dual AChE/BuChE inhibitor. Also, kinetic studies revealed that compound 4j exhibits mixed-type inhibition against both AChE and BuChE, with Ki values of 3.003 µM and 5.750 µM, respectively. Further, protein denaturation and proteinase inhibitory assays were used to test in vitro anti-inflammatory potential. It was found that compound 4o exhibited highest activity against protein denaturation (IC50 = 42.64 ± 0.19 µM) and proteinase inhibition (IC50 = 37.57 ± 0.19 µM) when compared to diclofenac. In addition, cell viability assay revealed that active compounds possess no cytotoxicity against N2a cell and RAW 264.7 macrophages. Finally, molecular docking experiments for AChE, BuChE, and COX-2 were conducted to better understand the binding modes of active compounds. Communicated by Ramaswamy H. Sarma


Introduction
Alzheimer's disease (AD) is a complicated age-related neurodegenerative condition marked by progressive memory, cognitive, and behavioural impairments. Several pathological hallmarks associated with AD are low acetylcholine (ACh) levels , agglomeration of b-amyloid (Ab) peptides into senile plaques (Brothers et al., 2018;Hampel et al., 2021); s-protein hyperphosphorylation into neurofibrillary tangles (Wegmann et al., 2021), oxidative stress (Rummel & Butterfield, 2021) and neuroinflammation (Guzman-Martinez et al., 2019;Kinney et al., 2018). Most of the medications used in clinical practice are based on the cholinergic theory i.e., inhibition of hydrolyses ACh into choline. Donepezil, rivastigmine and galantamine are only FDA approved cholinesterase inhibitors for the treatment of AD as they temporarily improve cognitive abilities in Alzheimer's patients by increasing the levels of acetylcholine (Colovic et al., 2013;Marucci et al., 2021). However, these AD drugs are known to impart a variety of adverse effects, including peripheral effects, hepatotoxicity, and gastrointestinal issues (Se Thoe et al., 2021;Sharma, 2019). Thus, it is a matter of priority to develop novel and efficacious pharmacophores that increase the levels of acetylcholine and serve as anti-Alzheimer agents. Acetylcholinesterase (AChE) and butyrylcholinesterase (BuChE) are two type cholinesterases. AChE is extensively found in the central nervous system (CNS), whereas BuChE is found in the peripheral nervous system (mostly in blood plasma) and aids in the hydrolysis of acetylcholine (ACh) (El-Sayed et al., 2020). Studies reveal that with progression of disease, AChE levels drop by up to 45% in Alzheimer's patients while BuChE levels rise by up to 90%, implying that BuChE substitutes for AChE's function (Giacobini, 2003;Greig et al., 2005). Furthermore, the crystallographic structure of ChE reveals that the primary binding sites are peripheral anionic site (PAS) and catalytic active site (CAS) (Kryger et al., 1999). The role of PAS is the inhibition of AChE activity by blocking ACh entry (Saini & Saxena, 2019) while CAS aids the breakdown of ACh (Pourshojaei et al., 2019). Thus developing ChE inhibitors that binds to both CAS and PAS of AChE as well as BuChE may be a promising lead in the development of anti-Alzheimer agents.
Many neurodegenerative disorders have neuroinflammation as a significant and unavoidable component. According to recent clinical research, neuroinflammation is hypothesized to play a role in the pathogenesis of Alzheimer's disease (Heppner et al., 2015). The presence of inflammatory markers in plasma and brain tissue led to the hypothesis that inflammation in the CNS has a role in the progression of neurodegeneration, particularly in Alzheimer's disease (Liu & Hong, 2003). It is reported that, in neurodegenerative illnesses, both neurons and glial cells express the cyclooxygenase-2 (COX-2) enzyme (Wyss-Coray & Rogers, 2012). Thus, compounds with anti-inflammatory potential may therefore address both anti-ChE inhibition and inflammatory response associated with it.
(Mumbai, India). Neuro2a cells (N2a) and RAW 264.7 macrophages were purchased from National Centre for Cell Sciences (NCCS), Pune, India. Absorbance was measured on Thermo Scientific Multiskan GO Microplate Spectrophotometer. Merck silica gel 60 F254 plates were used to monitor the reaction progress. Melting points were measured in open capillary tubes and are uncorrected. Perkin Elmer Frontier equipment with ATR was used to record the FTIR. 1 H NMR (300 MHz) and 13 C NMR (75 MHz) were recorded in CDCl 3 on Bruker AVANCE II using TMS as internal standard. ESI mass spectra were recorded on AB SCIEX 3200 QTRAP mass spectrometer. Elemental analysis (CHN) was performed on model EA300, Euro Vector, Italy.

In vitro cholinesterase activity
In vitro cholinesterase activity of the synthesized compounds was determined using Ellman's method (Ellman et al., 1961). Galantamine was used as a reference drug. Primarily, six different concentrations (0.625-20 mM) of standard and test drugs were prepared by diluting their stock solutions using sodium phosphate buffer (10 mM, pH 8.0). In 96-well plate, each well constituted a mixture of 20 mL of AChE (0.2 U/mL), 10 mL test compound, 150 mL sodium phosphate buffer and 10 mL of 5 mM DTNB solution. This mixture was incubated at 25 C for 15 min. Post incubation, 10 mL of 5 mM acetylthiocholine iodide (ATI) was added and the change in absorbance was determined spectrophotometrically at 412 nm using a plate reader with 60 sec intervals for 5 min. For control, an identical reaction mixture was prepared by substituting the sample solution with 10 mL sodium phosphate buffer. Percentage inhibition was calculated as per Eq. (1).

Inhibition activity %
IC 50 values were determined graphically from inhibition curves (log [Inhibitor] vs. % inhibition) in triplicate using GraphPad Prism 5.0. For BuChE assay, similar experiments were performed using BuChE (0.25 U/mL) and butyrylthiocholine chloride as substrate.

Kinetic analysis of ChE inhibition
For kinetic analysis purpose, the inhibition mechanism was evaluated spectrophotometrically at 412 nm for most active inhibitor 4j with slight modifications as previously described (Shaikh, Dhavan, Singh, et al., 2020). Herein, five different concentrations of substrate (acetylthiocholine/butyrylthiocholine; [S] ¼ 0.2, 0.25, 0.33, 0.5 and 1 mM) and three different concentrations of inhibitor ([I] ¼ 5, 10 and 20 lM) were used. Absorbance was determined and rate of reaction (V) was calculated. To evaluate the type of inhibition, Lineweaver-Burk (LB) plots were constructed by plotting 1/V vs. 1/[S] using GraphPad Prism 5.0. Further, inhibition constant (K i ) was determined using the slopes of the LB plots vs. inhibitor concentration [I]. The experiments were performed in triplicate.

Protein denaturation assay
In vitro protein denaturation was optimized according to reported method with minor modifications (S. F. . Briefly, in 3 mL reaction mixture 50 lL test sample and 450 lL BSA (5% w/v) was added. The reaction tubes were incubated at 37 C for 20 min and then heated at 57 C for 3 min. After cooling the tubes under tap water, 2.5 ml phosphate buffer saline (pH ¼ 6.3) was added to each tube and the absorbance was recorded spectrophotometrically at 660 nm. For control, an identical reaction mixture was prepared by substituting the sample solution with 50 mL phosphate buffer. Percentage protein denaturation inhibition was calculated as per Eq. (1).

Proteinase inhibitory assay
In vitro proteinase inhibitory assay was optimized according to reported method with minor modifications (S. F. . Initially, a concentration of 0.06 mg/ml trypsin, 2% casein, and 5% trichloroacetic acid at pH 7.6 was optimized. The reaction mixture of 3 ml constituted of 100 lL trypsin, 50 ll test samples and 350 ll of Tris-HCl buffer (25 mM, pH 7.4). The reaction mixture was incubated for 5 min at 37 C. Post incubation, 500 ll of casein was added and the mixture was further incubated for 20 min at 37 C. The reaction was quenched by addition of 2 ml 5% trichloroacetic acid and the cloudy suspension obtained was centrifuged for 10 min at 5000 rpm. The absorbance of supernatant solution was recorded at 280 nm. For control, an identical reaction mixture was prepared by substituting the sample solution with 50 mL Tris-HCl buffer. Percentage inhibition of the samples was calculated as per Eq. (1).

Cell viability.
(3-(4,5-dimethylthiazol-2-yl)-2,5diphenyl tetrazolium bromide (MTT) assay was performed in order to access the cell viability against N2a neuroblastoma cell and RAW 264.7 macrophages. 5 Â 10 3 cells/well were adhered in a 96-well plate and sustained in an incubator at 37 C under 5% CO 2 for 24 h. The cells were then treated with test compounds and incubated under similar conditions for 24 h. After incubation, MTT solution (20 mL, 0.5 mg/mL) was added to the wells and the plate was further incubated for 4 hours. Further, 100 mL of dimethyl sulfoxide (DMSO) was added to solubilise the formazan crystals. Thereafter, absorbance was recorded spectrophotometrically at 570 nm with a microplate reader and cell viability was determined by considering absorbance of the control cells as 100%. All the experiments were performed in triplicate.

Molecular docking
Molecular docking studies were performed using Glide 4.0 (Maestro, version 8.5, Schrodinger, LLC, 2008) software. The crystal structure of AChE (PDB ID: 1EVE), BuChE (PDB ID: 1P0I) and COX-2 (PDB ID: 4COX) were obtained from protein data bank. The crystal structure was optimized by removing water molecules, heteroatoms, and co-factors. Further, crystal structures were refined and the geometries of the conformations were optimized using the OPLS-2005 force field. Glide defaults for H-bonding was 2.5 Å whereas for amino acid residues were within a radius of 4 Å. Docking experiments were conducted using Glide XP docking following the standard protocol and parameters (Joshi et al., 2018).
Since no inhibitor is available for 1P0I, docking validation studies were carried out for PDB: 1EVE (complexed with donepezil) and 4COX (complexed with indomethacin) only. The inhibitors were extracted from their respective complex and re-docked ( Figure 2). The re-docked conformation (orange) of the inhibitor was aligned to the original conformation (green) and root mean square deviation (RMSD) was calculated. RMSD value obtained for 1EVE was found to be 1.163 while for 4COX it was found to be 0.465. This confirmed the accuracy of the docking protocol.

Chemistry
Antipyrine based a-aminophosphonates 4(a-o) were synthesized via one-pot solvent-free reaction of 4-aminoantipyrine, substituted aromatic aldehydes and diethyl phosphite at  80 C for 1 h (Scheme 1). Further the compounds were characterised using FTIR, 1 H-NMR, 13 C-NMR, Mass spectroscopy and elemental analyses (CHN). The characterization data is presented in experimental part.
At first the compounds were characterised using FTIR spectroscopy. Absorption bands around 3300 cm À1 , 1650 cm À1 , 1220 cm À1 and 1025 cm À1 can be assigned to NH, C ¼ O (amide), P ¼ O and P-O-C stretching respectively. 1 H NMR spectra exhibited a doublet around 5.0 ppm (J HP ¼ 21 Hz) which can be assigned to the proton at carbon which attached to phosphorus (CHP). Also, two singlet's around 2.1 ppm and 2.7 ppm can be ascribed to CH 3 and -NCH 3 of antipyrine. The group P-O-CH 2 -CH 3 displays two triplets around d ¼ 1.20-1.40 ppm and two multiplet around d ¼ 3.70-4.30 ppm for CH 3 and OCH 2 groups respectively. A peak around d ¼ 3.7 ppm can be allotted to NH proton.  Similarly, in 13 C-NMR a doublet is observed around d ¼ 57 ppm ( 1 J PC ¼ 150.7 Hz) for CHP carbon. Two singlet's around 10.5 ppm and 37 ppm can be ascribed to CH 3 and -NCH 3 of antipyrine. Also, P-O-CH 2 -CH 3 group displays two doublet around d ¼ 16-17 ppm and two doublet around d ¼ 62-63 ppm for CH 3 and OCH 2 groups respectively. The structure of the synthesized compounds was further elucidated using ESI-MS and C, H, N elemental analysis.

In vitro inhibition assay on cholinesterase
In vitro cholinesterase inhibitory activity of the synthesized derivatives 4(a-o) and standard drug galantamine were evaluated using Ellman's method (Ellman et al., 1961). The results are presented in Table 1 and are expressed in terms of IC 50 values. IC 50 values for AChE ranged between 0.46 to 6.67 mM and for BuChE IC 50 values were found to be in the range of 2.95 to 12.47 mM. Selectivity index for AChE was ranged between 2.04 to 6.45. The results display that synthesized derivatives possess comparatively more activity against AChE than BuChE. The most potent activity for AChE was observed in case of compound 4j (IC 50 ¼ 0.475 ± 0.12 mM) which was about 3-folds potent than galantamine (IC 50 ¼ 1.72 ± 0.11 mM). Other compounds that displayed better activity than galantamine are 4o (IC 50 ¼ 0.67 ± 0.13 mM), 4f (IC 50 ¼ 0.92 ± 0.14 mM), 4m (IC 50 ¼ 0.98 ± 0.13 mM), 4g (IC 50 ¼ 1.05 ± 0.11 mM) and 4e (IC 50 ¼ 1.14 ± 0.16 mM). However, in case of BuChE, compound 4j (IC 50 ¼ 2.95 ± 0.16 mM), 4m (IC 50 ¼ 3.28 ± 0.21 mM) and 4o (IC 50 ¼ 3.27 ± 0.14 mM) Data are means ± SD of three independent experiments and % cell viability was calculated relative to untreated control. Table 2. Detailed comparative interaction analysis of compound 4j, 4m and 4o with their standard inhibitor and co-crystallised ligand in the active site of AChE, BuChE and COX-2. displayed better activity than galantamine (IC 50 ¼ 3.62 ± 0.11 mM). From structure activity point of view, it is observed that AChE activity for mono-substituted aldehyde at para-position was found in the order: 4-N(Me) 2 > 4-OMe > 4-F > 4-Cl > 4-Me > 4-H. For methoxy substituent, the AChE activity follows the order: di-substituted > mono-substituted > tri-substituted. In case of meta-substitution, AChE activity followed the order: 3-OH > 3-NO 2 > 3-F. Also, AChE activity for hetero-aldehyde followed the order: thiophene > furan. However, in case of BuChE, no exact correlation between the nature of the substituent and BuChE activity can be interpreted. Compound 4j (IC 50 ¼ 2.95 ± 0.16 mM), 4m (IC 50 ¼ 3.28 ± 0.21 mM) and 4o (IC 50 ¼ 3.27 ± 0.14 mM) exhibited comparative activity as that of galantamine (4m (IC 50 ¼ 3.62 ± 0.11 mM). It is noteworthy that compound 4j, 4m and 4o exhibits potential AChE as well as BuChE activity suggesting that they act as dual AChE/BuChE inhibitor.

Kinetic analysis of AChE and BuChE inhibition
Kinetic studies of AChE and BuChE were evaluated in order to verify the mechanism of enzyme inhibition (Figure 3). Inhibition mechanism for the most active compound i.e., 4j was determined graphically by constructing a Lineweaver-Burk plot of 1/V versus 1/ [S]. The results demonstrated that compound 4j undergoes a mixed-type inhibition i.e., it is able to bind to PAS and CAS of both the enzymes (AChE and BuChE). Further, secondary plots of slope versus inhibitor concentration [I] were plotted to determine the inhibition constant (K i ). K i for AChE and BuChE was found to be 3.003 mM and 5.750 mM respectively.
3.2.3. In vitro anti-inflammatory activity 3.2.3.1. Protein denaturation assay. One of the major causes of inflammation is protein denaturation which usually results when proteins lose their secondary and/or tertiary structure on exposure to heat or harsh chemicals (Mengji et al., 2015). In the process of denaturation, a protein transforms itself from soluble form to an insoluble form leading to loss of its biological role (Cohn, 1925). Hence, anti-inflammatory activities of the synthesized derivatives against inhibition of protein denaturation were studied. The results are presented in Table 1 in terms of IC 50 values. The results suggest that all the tested compounds have some anti-inflammatory potential as they can inhibit the denaturation of albumin. Compound 4o (IC 50 ¼ 42.64 ± 0.19 mM) and 4j (IC 50 ¼ 44.15 ± 0.16 mM) exhibited better activity when compared to standard drug diclofenac (IC 50 ¼ 46.29 ± 0.12 mM).
3.2.3.2. Proteinase inhibitory assay. Neutrophils are found in lysosomes and play an important role in the pathogenesis of inflammation as they are rich in serine proteinase (Bacha et al., 2017). During inflammatory reactions, tissue damage is caused by these leukocyte proteinases. Anti-inflammatory property of a drug has been linked to inhibition of these proteinases (Rastogi et al., 2018). Thus, inhibition of proteinase activity by the synthesized compounds was investigated in order to evaluate their anti-inflammatory ability. The results are presented in Table 1 in terms of IC 50 values. From the results, it is evident that suggest that all the tested compounds have ability to inhibit proteinase in comparison with control. Compound 4o (IC 50 ¼ 37.57 ± 0.19 mM and 4j (IC 50 ¼ 41.25 ± 0.18 mM) exhibited better activity when compared to standard drug diclofenac (IC 50 ¼ 45.76 ± 0.14 mM) while other derivatives exhibited moderate activity.

Cell viability assay
MTT assay was performed in order to access the cell viability against N2a neuroblastoma cell and RAW 264.7 macrophages. For this purpose, the most active compounds i.e., 4j, 4m and 4o were evaluated for their cytotoxicity (cell death) and the results were compared with control. In case of N2a cells, cell viability was determined by exposing cells to higher concentration of test compounds (20 and 40 lM) for 24 h. The results are displayed in Figure 4(a). The results suggest that test compounds do not impart any toxicity to N2a cells. In case of RAW 264.7 cells, cell viability was determined by exposing cells to higher concentration of test compounds (50 and 100 mg/ ml) for 24 h. The results are displayed in Figure 4(b). From, the results it is evident that no cytotoxicity was observed for the test compounds.

Molecular docking studies
Molecular docking studies were performed using Glide 4.0 (Maestro, version 8.5, Schrodinger). Glide identifies favourable interactions and predicts glide score, a scoring system that calculates the ligands' binding free energy with the protein. Molecular docking results are delineated in Table 2.

Docking against cholinesterases
In case of AChE, docking experiments were conducted using PDB:1EVE, due to its availability of crystal structure and the established understanding of ligand interactions (de Lima et al., 2019;Gupta et al., 2011). However, in case of BuChE, human BuChE (1P0I) was used in the docking study because the crystal structure of BuChE from equine serum has not been reported and also the sequence of equine BuChE is quite similar to that of human BuChE (Luo et al., 2013). Figure 5 represents detailed comparative analysis of active site of AChE and BuChE. In the active site of AChE, amino acid residues are distributed in different binding sites such as PAS (Tyr70, Asp72, Tyr121, Trp279 and Tyr334), anionic site (Trp84, Glu199, Phe330 and Phe331), esteratic subsite (Ser200, Glu327 and His440) and acyl pocket (Phe288 and Phe290) (Dvir et al., 2010). Anionic site, esteratic subsite and acyl pocket together constitute CAS. BuChE and AChE have a 53% sequence homology, indicating that they are quite similar (Harel et al., 1992;Su arez et al., 2006). The difference between BuChE and AChE exists in the acyl binding pocket, where two smaller amino acid residues (Leu286 and Val288) substitute the larger residues (Phe288 and Phe290) in AChE, allowing the accommodation of bulky substrate in the BuChE cavity (Brus et al., 2014;Kumar et al., 2018). In the PAS of BuChE, only counterparts of AChE present are of Asp72 and Tyr334 (i.e., Asp70 and Tyr332) and do not own counterparts of Tyr70, Tyr121 and Trp279 residues. Also, in the anionic site of BuChE only Trp82 residue is conserved whereas Phe330 residue is replaced by Ala328 residue (Bajda et al., 2013). This lack of some important residues influences the inhibitory affinity of BuChE in comparison to AChE. Figure 6 represents 3D docking poses of compound 4j, 4m, 4o and galantamine in the active site of AChE. From the Figure 6, it is evident that the docked compounds are able to fit well in active site as that of galantamine and interact with important amino acid residues. The reference drug i.e., galantamine bears three hydrophobic interaction with Phe330 (CAS), Phe331 (CAS), Tyr334 (PAS) while two hydrogen bonding interactions with Asp72 (PAS), Tyr130 (CAS). On the other hand, the most active compound i.e., 4j displays two p-p stacking interaction with Tyr70 (PAS) and Phe330 (CAS) residues while four established hydrogen bonding interaction with Asp72 (PAS), Tyr121 (PAS), Phe288 (CAS) and Arg289 (CAS). These results suggest that compound 4j has better binding potential than the reference drug. Also, the above findings corroborate with the mixed type inhibition obtained in kinetic studies suggesting that compound 4j binds to both CAS and PAS of AChE. Additionally, the interactions for the synthesized derivatives were compared with the co-crystallized ligand i.e., donepezil (Table 2) and were found to be similar (see supporting information Figure S1).
Similarly, Figure 7 represents 3D docking poses of compound 4j, 4m, 4o and galantamine in the active site of BuChE. Compound 4j exhibits two p-p stacking interaction with Trp231 (CAS) and Phe329 (CAS) while one hydrogen bonding interaction with Glu197 residue. Galantamine possess only two hydrogen bonding interactions with Asp70 and His438 residues. This attributes to the fact the 4j possess better binding potential than galantamine.

Docking against COX-2
Aforementioned that COX-2 is expressed in neuroinflammation, thus docking studies were carried out using COX-2 enzyme (PDB: 4COX). Active site of COX-2 enzyme comprises of residues like Arg120, Gln192, Tyr355, Trp387, Glu524 and Ser530 (Fatahala et al., 2017). However, Tyr385 and Ser530 play a pivotal role in the enzyme activity. Tyr385 limits COX enzyme activity by causing radical production and hydrogen abstraction, whereas hydrogen bonding with Ser530 causes irreversible inhibition (Garavito & Mulichak, 2003). Thus, molecular docking studies for the most active inhibitors 4j, 4m and 4o were performed against COX-2. The results are displayed in Table 2. Figure 8 shows 3D docking poses of compound 4j, 4m, 4o and diclofenac. The results suggest that the docked entities exhibit similar docking features to diclofenac as they bind in the same cavity. Diclofenac exhibits two hydrophobic interactions with Tyr385 and Trp387 while one hydrogen bonding interaction with Ser530. Compound 4j exhibits a p-p stacking interaction as well as H-bonding interaction with Tyr355 and Leu352 residues respectively. Similarly, compound 4m shows a p-p stacking interaction with Tyr385 and H-bonding interaction with Arg120. Compound 4o displays two p-p stacking interaction with Tyr385 and Trp387 residues. Further, the interactions for the synthesized derivatives were compared with the co-crystallized ligand i.e., indomethacin ( Table 2) and were found to be similar (see supporting information Figure S2).

Conclusion
In conclusion, a series of antipyrine based a-aminophosphonates 4(a-o) were synthesized and characterized. The synthesized derivatives were subjected for anti-cholinesterase and anti-inflammatory activity. The results display that synthesized derivatives possess better AChE activity than BuChE. Compound 4j exhibited highest inhibition against AChE (IC 50 ¼ 0.475 ± 0.12 mM) as well as BuChE (IC 50 ¼ 2.95 ± 0.16 mM) implying that it acts as a dual AChE/BuChE inhibitor. Further, kinetic analysis inferred that compound 4j undergoes a mixed-type inhibition against both AChE and BuChE with K i value 3.003 mM and 5.750 mM respectively. In vitro anti-inflammatory activity was carried out using protein denaturation assay and proteinase inhibitory assay. Also, cell viability assay was evaluated against N2a cell and RAW 264.7 macrophages. Finally, molecular docking studies were carried out for AChE, BuChE and COX-2 to understand the binding mode of the active compounds.