Synthesis, biological evaluation and molecular docking studies of new 2- ethoxy-4-{[3-alkyl(aryl)-4,5-dihydro-1H-1,2,4-triazol-5-on-4-yl]- azomethine}-phenyl benzenesulfonate derivatives on human aldose reductase enzyme

A series of 2-ethoxy-4-{[3-alkyl(aryl)-4,5-dihydro-1H-1,2,4-triazol-5-on-4-yl]-azomethine}phenyl benzenesulfonates (3) were synthesized from the reactions of 3-alkyl(aryl)-4-amino4,5-dihydro-1H-1,2,4-triazol-5-ones (1) with 2-ethoxy-4-formyl-phenyl benzenesulfonate (2). N-acetyl derivatives (4) of compounds 3 were also obtained. Then, the compounds 3 have been treated with morpholine and 2,6-dimethylmorpholine in the presence of formaldehyde to synthesize 2-ethoxy-4-{[1-(morpholine-4-yl-methyl)-3-alkyl(aryl)-4,5-dihydro-1H-1,2,4triazol-5-on-4-yl]-azomethine}-phenyl benzenesulfonates (5) and 2-ethoxy-4-{[1-(2,6dimethylmorpholine-4-yl-methyl)-3-alkyl(aryl)-4,5-dihydro-1H-1,2,4-triazol-5-on-4-yl]azomethine}-phenyl benzenesulfonates (6), respectively. The structures of twenty-six new compounds were identified by using elemental analysis, IR, H NMR, C NMR, and MS spectral data. In addition, in vitro antibacterial activities of the new compounds were evaluated against six bacteria such as Escherichia coli, Pseudomonas aeruginosa, Staphylococcus aureus, Bacillus subtilis, Bacillus cereus and Klepsiella pneumonia according to agar well diffusion method. Furthermore, in order to determine the possible antidiabetic properties of the synthesized 1,2,4-triazole derivatives, inhibition effects on the AR enzyme were investigated and molecular docking studies were carried out to determine the receptorligand interactions of these compounds. IC50 values of triazole-derived compounds (6a, 6b, 6d-g) against AR enzyme were determined as 0.95 μM, 0.75 μM, 1.83 μM, 0.62 μM, 1.05 μM, 1.06 μM, respectively. Considering the docking scores and binding energies obtained docking studies, it has been shown that molecules fit very well to the active site of the AR enzyme.

The classical Mannich reaction, a three-component condensation between structurally diverse substrates containing at least one active hydrogen atom, an aldehyde component and an amine reagent leads to a class of compounds known as Mannich bases (Roman 2015). Mannich bases have applications in the pharmaceutical field and in other industries, such as the petroleum, the cosmetics, the dyes and the food industries, etc. The principal advantage of the Mannich reaction is that it enables two different molecules to be bonded together in one step (Tramontini and Angiolini 1994). Mannich bases acquired from 1,2,4-triazole derivatives are reported to possess biological activities including; antifungal, antioxidant, antilipase and antibacterial properties (Manap et al. 2020;Wang et al. 2017;Ceylan 2016).
Diabetes mellitus (DM), the most common chronic disease known to date, is a disorder affecting approximately 463 million people worldwide, and by 2045 the number of diabetic patients is expected to increase to 700 million (Saeedi et al. 2019). Numerous studies have shown that diabetes mellitus can cause various complications such as cardiovascular complications, diabetic nephropathy, liver complications, neuropathy, and cataract, which are the main causes of morbidity and mortality (Meng et al. 2019). Hyperglycemia is a condition characterized by diabetes and plays an important role in the development and progression of these complications resulting from irreversible long-term damage to biological macromolecules as well as acute and reversible changes in cellular metabolism (Ottana et al. 2011). Under hyperglycemic conditions, autooxidation reactions of glucose promoting intermolecular crosslinking and lipid oxidation produce reactive oxygen species (ROS) that can damage tissues, cellular membranes, and biomolecules (Brownlee 2005). In addition, highly reactive dicarbonyl compounds produced during glucose auto-oxidation reactions lead to the glycation of macromolecules resulting in the synthesis of advanced glycation end products (AGEs) (Jay et al. 2006).
In the case of hyperglycemia, a significant increase in polyol pathway activity increases both cellular osmotic stress and oxidative stress, thus contributing significantly to diabetes-related tissue damage, particularly in insulin-independent tissues of glucose uptake (Giacco and Brownlee 2010). Increased glucose flux along the polyol pathway triggers a range of metabolic events that can result in tissue and vascular damage, decrease in blood flow and decrease nerve conduction rate, thereby critically contributing to the etiology of 3 DM-related chronic complications (Maccari and Ottanà 2015). Especially in the lens, the accumulation of sorbitol in the cell leads to swelling, membrane permeability changes and osmotic stress, thereby promoting cataractogenesis (Brownlee 2001). In addition, NADPH depletion due to activation of the polyol pathway in hyperglycemic conditions disrupt the activity of NADPH-dependent enzymes such as glutathione reductase and indirectly contributes to the production of intracellular reactive oxygen species (Evans et al. 2002). However, the changing NADH/NAD + ratio, during the second step of the polyol pathway (conversion of sorbitol to fructose by SDH enzyme), increases the redox imbalance, causing many metabolic and signal pathways to be affected (Williamson et al. 1993).
Aldose reductase (EC1.1.1.21) (ALR2), an enzyme detected in the lens, brain, skeletal muscle, liver, and kidney tissue, is an aldo-keto reductase that catalyzes the reduction of glucose to sorbitol using the NADPH cofactor (Maccari et al. 2008). The AR enzyme has been extensively investigated as an enzyme that has been critically involved in the emergence and progression of many pathologies associated with diabetes mellitus, especially in the past four decades. It is clear that the way to onset and delay diabetic complications is to block the activity of the polyol pathway through the inhibition of aldose reductase enzyme. However, there is no worldwide commercially available antidiabetic drug that inhibits the aldose reductase enzyme, yet (except for some eastern countries) (Balestri et al. 2017). So, in our study, the inhibition effects of newly-synthesized triazole compounds on the recombinant human aldose reductase enzyme were investigated and enzyme-inhibitor interactions are computed by molecular docking studies.

Antimicrobial Activity
The microbiological results are given in Table 1. Compounds 3 and 6 against to Bacillus subtilis and compounds 5 against to Bacillus cereus and Escherichia coli did not display any antimicrobial activity. The active compounds are emphasized by using bold italic characters in the table. The highest zone diameter was obtained from the compound 3f against Bacillus cereus and compounds 6a and 6f against Escherichia coli. The compounds 4 did not display any antimicrobial activity against to all tested microorganisms.

In Vitro Inhibition and Molecular Docking Studies
In the present study, the inhibition effects of newly synthesized compounds 6a, 6b, 6d-g on recombinant human aldose reductase enzyme were investigated. IC50 and Ki values of newly synthesized triazole compounds were calculated, and inhibition types were determined from 6 Ki plots. IC50 values of compounds 6a, 6b, 6d-g were determined as 0.95 µM, 0.75 µM, 1.83 µM, 0.62 µM, 1.05 µM, 1.06 µM, respectively. Ki values were calculated as 0,42±0,09 µM, 0,74±0,49 µM, 1,43±0,41 µM, 0,27±0,02 µM, 0,68±0,33 µM, 2,04±0,35 µM, respectively. The inhibition type of the compounds 6a and 6f were competitive, the inhibition type of the compound 6e was uncompetitive, and the inhibition type of the other compounds was determined as non-competitive (Table 2). Molecular docking studies were also conducted to support and verify experimental inhibition studies. Molecular docking studies showed that compounds 6a, 6e and 6f were the compounds that best interacted with the active site of the AR enzyme, according to docking scores and binding energies. It has been shown that the molecular docking results are highly correlated with the experimental results and the inhibition mechanism is valid. One of the most widely used docking validations is the re-docking method (Yuriev et al. 2011). In this method, the ligand is extracted from the receptor and re-dock to the respective protein. In this study, co-crystallized ligand tolrestat was extracted from the receptor and re-docked into the active site of the AR enzyme. After docking, the best pose of the tolrestat was aligned with the co-crystalized ligand and RMSD (Root Mean Square Deviation) of the ligand was calculated. RMSD value less than 2 Å proves the accuracy of the docking protocol (Wang et al. 2003). When the best pose of the tolrestat conformation superimposed with the co-crystallized ligand, RMSD value determined as 0,9812 Å (Fig. 3). In recent years, triazoles and their heterocyclic derivatives have attracted great attention due to their potent pharmacological properties. Numerous triazole compounds have often been used as clinical drugs or candidates for the treatment of various types of diseases, showing wide potential as medicinal agents (Zhou and Wang 2012). It has been shown in many studies that triazole derivative compounds have antioxidant (Barbuceanu et al. 2014), anticancer (Tyagi et al. 2015), antibacterial (Almajan et al. 2009), antitubercular (Shaikh et al. 2016), analgesic (Vijesh et al. 2013), antifungal (Wu et al. 2018), anti-inflammatory (Khan et al. 2020), anticonvulsant (Sari et al. 2018), and antiviral (Chen et al. 2019) properties. Although the antidiabetic properties of triazole derivatives are known, there is no experimental study on the synthesis of triazole compounds targeting the aldose reductase enzyme. Therefore, in this study, we synthesized six new triazole-derived compounds (6a, 6b, 6d-g) and examined the inhibition effects of these compounds on the diabetes mellitus related enzyme aldose reductase. IC50 values of compounds 6a, 6b, 6d-g determined as 0.95 µM, 0.75 µM, 1.83 µM, 0.62 µM, 1.05 µM, 1.06 µM, respectively. The inhibition results show that all compounds synthesized inhibited the aldose reductase enzyme at micromolar and submicromolar concentrations. According to the inhibition results, the best potential inhibitor for AR enzyme was compound 6e (IC50= 0,62 µM), while compound 6d (IC50= 1,83 µM) showed the lowest inhibition effect. In addition to determining the IC50 values of the synthesized compounds, Ki values, which are an expression of the inhibitors' affinity for the enzyme, were also determined. Ki values of the synthesized triazole derivative compounds 6a, 6b, 6d-g were determined as 0,42±0,09 µM, 0,74±0,49 µM, 1,43±0,41 µM, 0,27±0,02 µM, 0,68±0,33 µM, 2,04±0,35 µM, respectively. According to these results, it is clear that the inhibitor with the highest affinity for the AR enzyme is compound 6e. It is seen that the determined Ki constants are correlated with the inhibition results.
Molecular docking studies were also conducted to theoretically verify the experimentally obtained inhibition results and to elucidate the inhibition mechanisms by determining the inhibitor-receptor interactions. Therefore, the newly synthesized triazole derivatives were docked against aldose reductase enzyme (PDB code: 2FZD) (AR-tolrestat adduct). With docking studies, docking score, XP GScore (Extra precision glide score), glide evdw (Van der Waals energy), glide ecoul (coulomb energy), glide energy (Model energy) and glide emodel (Modified Coulomb-van der Waals interaction energy) values of the compounds were determined (Table 3). Considering the docking scores and binding energies, it is observed that the compound that interacts best with the active site of the AR enzyme is 6a, which has the lowest IC50 value with competitive inhibition type. The compound 6a made hydrogen bonding with the amino acid residues TYR48 and CYS298 and pi-pi stacking with the amino acid residues HIS110 and TYR209 in the active pocket of the AR enzyme ( Fig.  2A). It has been confirmed that compound 6e, which shows the best inhibitory effect experimentally, is one of the best inhibitor candidates, considering the docking score and binding energies.
In order to emphasize how important the results obtained from docking studies for the synthesized triazole derivative compounds, docking studies were also performed for wellknown AR enzyme inhibitors tolrestat, sorbinil, and zopolrestat. It shows that the compounds 6a, 6b, 6e and 6f, which have very good inhibitory effects against the AR, interact better with the catalytic active site of the AR enzyme than both sorbinil and zopolrestat (Table 3).

Conclusion
In the present study, new 1,2,4-triazole derivatives (3-6) were designed and synthesized. Their structures were identified using elemental analysis, IR, 1 H NMR, 13 C NMR and MS spectral data. The target compounds were also investigated for their antimicrobial potential. In addition, the inhibition effects of the six 1,2,4-triazole compounds on the AR enzyme were examined and docking studies were carried out to determine the inhibitor-receptor interactions. Inhibition studies have shown that all of the compounds investigated, especially compound 6e, are strong potential AR enzyme inhibitors. Docking studies of the investigated compounds against the AR enzyme show that the compounds fitted in the active site of the enzyme very well, compared to some well-known standard AR inhibitors and confirmed the experimental inhibition results and mechanism of inhibition.

Chemistry
Chemical reagents used in this paper were bought from Merck AG, Aldrich and Fluka. Recombinant human aldose reductase enzyme, β-nicotinamide adenin dinucleotide phosphate (β-NADPH), DL-Glyceraldehyde, and sodium phosphate were purchased from Sigma-Aldrich Chemie GmbH (Taufkirchen, Germany). Melting points were taken using an Electrothermal Melting-point Apparatus in an open capillary tube and were not corrected. The infrared spectra were recorded on a Perkin Elmer Instruments Spectrum One FT-IR spectrophotometer. 1 H and 13 C NMR spectra were determined in deuterated dimethyl sulfoxide with TMS as internal standard using a Bruker Ultrashield spectrophotometer at 400 MHz and 100 MHz, respectively. Elemental analyses were carried out on a LECO, CHNS-932 for C, H, and N. Electrospray ionisation mass spectrometry (ESI-MS) was performed on a TSQ Quantum Access Max Triple Stage Quadrupole Mass Spectrometer.

Antimicrobial Activity
All bacterial and yeast strains were obtained from the company of Microbiological Environmental Protection Laboratories (France) and were as follows: Bacillus Substilis (ATCC 11774), Bacillus Cereus (ATCC 11778), Staphylococcus aureus (ATCC 6538), Escherichia coli (ATCC 25922), Pseudomonas aeruginosa (ATCC 27853), Klebsiella pneumonia (ATCC 4352). Simple susceptibility screening test using agar well diffusion method was used (Perez et al. 1990;Ahmad et al. 1998). All the newly synthesized compounds were weighed and dissolved in dimethylsulphoxide (DMSO) to prepare extract stock solution of 1 mg/ml.
Each microorganism was suspended in Mueller-Hinton Broth and diluted to 106 colony forming unit (cfu) per ml. They were "flood-inoculated" onto the surface of Mueller Hinton Agar and then dried. Five-millimeter diameter wells were cut from the agar using a sterile cork-borer, and 250-5000 μg/50 μl of the chemical substances were delivered into the wells. The plates were incubated for 18 h at 35 °C. Antimicrobial activity was evaluated by measuring the zone of inhibition against the test organism. Ampicillin, neomycin and streptomycin were standard antibacterial and antifungal agents, DMSO was used as solved control.

Aldose Reductase Activity Assay and Inhibition Studies
In vitro activity analysis of aldose reductase enzyme inhibition was performed according to the method described by Cerelli (Cerelli et al. 1986). The assay mixture containing 1 M sodium phosphate buffer (pH 5.5), 0.11 mM NADPH, enzyme solution, and distilled water was prepared. The reaction was started by adding 4.7 mM DL-Glyceraldehyde to the reaction medium, and the change in absorbance at 340 nm due to NADPH oxidation was measured with the aid of a UV-Visible Spectrophotometer. In order to determine the inhibition effects of the synthesized triazole derivatives, inhibitors were added to the reaction medium at different concentrations and activity measurements were performed. Control activity without inhibitor was considered as 100% and inhibitor concentrations that halve enzyme activity were determined as IC50. IC50 values were calculated from the equation of the graph obtained by calculating the activity% of the enzyme measured in different inhibitor concentrations. In order to calculate Ki values of triazole derivatives showing the inhibitory effect on aldose reductase enzyme, activity values were determined with 3 different inhibitor concentrations 22 against 5 different substrate concentrations. Lineweaver-Burk plots of each compound were drawn from the activity values obtained and Ki values and inhibition types were determined from graphics equations (Alim et al. 2017;Taslimi et al. 2018).

Molecular Docking Studies
Molecular docking method was used to determine ligand-receptor interactions. For this purpose, high resolution X-ray crystal structure of aldose reductase enzyme was retrieved from RCSB Protein Data Bank (PDB code: 2FZD, Resolution: 1.08 Å ). The obtained structure from PDB was pre-processed using Protein Preparation Wizard (Sastry et al. 2013) in Schrödinger Maestro software (Maestro Version 12.4.072, Release 2020-2). Protein Preparation Wizard includes the proper assignment of bond orders and ionization states, simplifying multimeric structures, adding and optimizing hydrogen bonds, converting selenomethionines into methionines, locating and removing unnecessary water molecules, the assignment of partial charges, forming disulfide bonds, aligning and closing terminal amides. In addition, the addition of missing atoms and side-chain residues accomplished with the Prime module. Amino acids were ionized at physiological pH using Propka module. Water molecules at a distance of 3 Å from the receptor were removed. Energy minimization was done using OPLS-2005 force field. 3D structures of the ligands were obtained with the help of LigPrep program. Epik module and OPLS-2005 force field were used to obtain correct molecular geometries and protonation states of the ligands at pH 7.0 ± 2.0. Using the Receptor Grid Generation platform, the possible binding site of the ligands to be docked was determined. Then, using Glide, the prepared ligands were docked to the target enzyme. Docking calculations were performed by keeping the ligand flexible using the Glide XP (extra precision) algorithm. In order to verify the docking procedure, docking validation was also performed (Yuriev et al. 2011). For this purpose, the ligand bound to the receptor was extracted and re-docked to the protein. After docking, the best pose of ligand was aligned with the co-crystallized ligand and the RMSD (Root Mean Square Deviation) value was calculated. RMSD value less than 2 Å proves the accuracy of the docking protocol (Bal et al. 2020).