A computational investigation of galactopyranoside esters as antimicrobial agents through antiviral, molecular docking, molecular dynamics, pharmacokinetics, and bioactivity prediction

Abstract One of the most common viral infections worldwide is the Human Papilloma Virus (HPV) which has been linked to cancer and other diseases in many countries. Monosaccharide esters are significant in the field of carbohydrate chemistry because they are efficient in the synthesis of pharmacologically active compounds. Therefore, the present study aimed to perform thermodynamic, molecular docking and molecular dynamics study of a series of previously designed monosaccharaides, methyl β-d-galactopyranoside (MGP, 1) esters (2–10) with along with their physicochemical and pharmacokinetic properties. We have optimized the MGP esters employing the DFT study at the B3LYP/6-311 + G (d,p) level of theory. The subsequent analysis also investigated the electronic energies, enthalpies, entropies, polarizability, and natural bond orbital (NBO) of these modified esters. Then, MGP esters were docked into CTX-M-15 extended-spectrum beta-lactamase from Escherichia coli (PDB: 4HBT) and E2 DNA-binding domain from human papillomavirus type 31 (PDB: 1A7G), and the results revealed that most of the esters can efficiently bind to the target. Desmond was used to doing molecular dynamics simulations at 200 ns in addition to molecular docking to look at the binding conformational stability of the protein–ligand complex. Based on RMSD and RMSF, it was determined that the stability of the protein-ligand combination was maintained during the whole 200 ns simulations for all compounds. Finally, a pharmacokinetic study suggests that modified esters of MGP exhibited better pharmacokinetic characteristics and were less hazardous than the parent drug. This work demonstrated that potential MGP esters can efficiently bind to 4HBT and 1A7G proteins and opened avenues for the development of newer antimicrobial agents that can target dangerous pathogens. Communicated by Ramaswamy H. Sarma


Introduction
Monosaccharides are key natural molecules with several biological processes.Medicinal experts and researchers around the world are searching for new effective antimicrobial and antiviral agents against different pathogenic organisms.For these reasons, synthesizing new chemicals and examining their antimicrobial activity are the best ways to develop efficient antimicrobial and antiviral agents.The significant importance of carbohydrates in biological systems, such as bacterial and viral infections, cell development and proliferation, cellcell communication, and immunological response, makes them a very appealing topic for scientists (Chen & Fukuda, 2006;Nogueira et al., 2009).In addition to providing metabolic energy, they also help to optimize cell-cell communication and other vital functions (Peter & Werz, 2007;Varki, 1993).It was found from the literature survey that many biologically active molecules also have aromatic, heteroaromatic, and acyl substituents (Alam et al., 2021;Arifuzzaman et al., 2018;Bulbul et al., 2021;Chowdhury et al., 2016;Kawsar et al., 2015Kawsar et al., , 2018;;Maowa et al., 2021;Rana et al., 2020).Benzene, substituted benzene, and also heteroatoms and halogen-containing substituents are known for enhancing the biological activity of the parent compound (Alam et al., 2021;Devi et al., 2019;Rana et al., 2021;Yasmin et al., 2021).The incorporation of an active nucleus into an inactive compound will enhance the potential biological effects (Kawsar & Kumer, 2021;Islam et al., 2019).Furthermore, selective acylation of carbohydrates, as well as evaluation of microbial activities (Kabir et al., 2009;Kawsar et al., 2014) revealed that in many cases, the combination of two or more heteroaromatic nuclei and acyl groups improves the biological activity manifold than its parent nucleus (Kawsar et al., 2020).Monosaccharide esters have a broad spectrum of activity against Gram-negative and Grampositive organisms such as Escherichia coli, Bacillus subtilis, Salmonella typhi, and Staphylococcus aureus (Islam et al., 2019;Kabir et al., 2009;Kawsar et al., 2014Kawsar et al., , 2020)).Recently, it was proved that some monosaccharide analogs were found to possess great inhibition against cancer cells (Kawsar et al., 2020).On the other hand, it was revealed that the modified hydroxyl (-OH) group of nucleoside and monosaccharide structures exhibit new potent antimicrobial and antiviral activities (Bulbul et al., 2021;Farhana et al., 2021;Kawsar et al., 2012;Kawsar & Hossain, 2020;Maowa et al., 2021).However, Gram-negative bacteria develop resistance through the production of b-lactamase (Breijyeh et al., 2020).Therefore, it becomes most important to work on this bacterial resistance by developing newer antibacterial agents.Also, one of the most common viral infections worldwide is the Human Papilloma Virus (HPV), which has been linked to cancer and other diseases in many countries (Chan et al., 2019).The carbohydrates derivatives can form a large group of natural proteins, peptides, and synthetic compounds that can interact significantly with glycosylated proteins.Carbohydrate derivatives can be synthesized and also isolated from various species.Monosaccharides like mannose, glucose, galactose, and residues found in the backbone of N-glycan structures all interact differently with each carbohydrate-binding substance.Because many enveloped viruses, such as HIV, HCV, DENV, and HPV, are glycosylated at the viral surface, carbohydrate derivatives could interact with the virus's glycosylated envelope and prevent viral entry into the host cell (Alen & Schols, 2012;Sarkar et al., 2021).Previously, several carbohydrate-binding agents isolated from plants and algae, specifically binding mannose and N-acetyl glucosamine, were shown to have antiviral activity against HIV and HCV (Balzarini et al., 2007).Since the N-glycans on the viral envelope are directly bound by the carbohydrate derivatives, they inhibit viral entry.In addition, they indirectly compel the virus to delete its N-glycans and lose the capacity to enter the dendritic cell (Balzarini et al., 1991).In the presented research, the MGP-hydroxyl groups were acylated by some aromatic and heterocycle acyl substituents (Amin et al., 2022;Islam et al., 2022).It is reported the MGP is a derivative form carbohydrate including the group of sugars.Many scientists have been reported that this derivative consists a wide variety of antifungal, antibacterial, anticancer properties, antiviral and SARS-CoV-2 (Amin et al., 2021).Due to have the vast attention as potential molecules for drug design, the MGP and its esters have selected to evaluate the biological significance against various human pathogens using computational tools (Kawsar et al., 2022).It is reiterated that the MGP esters have reported antibacterial and antifungal properties so that this study conveys the various chain effects in replacing the acyl group (Kabir et al., 2022;Kawsar et al., 2023).The new modified esters were investigated to determine their thermal and electrical stability and physicochemical properties using quantum mechanical methods.Keeping the above features in consideration, molecular docking studies were performed to investigate the possible binding mechanism of MGP esters with native CTX-M-15 extended-spectrum beta-lactamase from E. coli (PDB: 4HBT) and the E2-DNA-binding domain from human papillomavirus type 31 (PDB: 1A7G).To compare their absorption, metabolism, and toxicity, pharmacokinetic enumeration was lastly carried out.Understanding the thermochemical, atomic partial charge, binding mechanism, physicochemical, and pharmacokinetic features of MGP (1) and its esters was the main goal of our work (2-10).Several computational studies, including molecular docking and molecular dynamics simulation, were used to examine the binding interactions between a prospective lead compound and the target in this work.The goal of the research is to uncover possible leads and speed up the development of new antimicrobial drugs.

Geometry optimization
The Quantum mechanical methods are usually utilized to calculate thermal, molecule orbital, and molecular electrostatic characteristics by computational application.All designed analogs were subjected to geometry optimization and subsequent modification using the Gaussian 09 program (Frisch et al., 2009).To optimize and predict their thermal and physicochemical properties, the DFT with Beck's (B) (Becke, 1988) three-parameter hybrid model and Yang, Lee, and Parr's (LYP) (Lee et al., 1988) correlation functional under a 6-311 þ G (d,p) basis set was used.For each compound, partial atomic charge, polarizability, heat capacity, total energy, enthalpy, and entropy were calculated from the optimized files.

Antiviral activities prediction
Antiviral molecules (AVMs) are used in antimicrobial medication to treat viral infections by reducing viral pathogen growth inside the host cell.The internet software (http:// crdd.osdd.net/servers/avcpred) is used to calculate antiviral activity.For predictions, the SD file format of the MGP esters was inputted.The evaluation was carried out to aid in the development of antiviral therapeutics as well as to recommend the best inhibitory MGP esters for future research.

Preparation of protein and molecular docking
The 3D structures of 4HBT (Lahiri et al., 2013) and 1a7g (Bussiere et al., 1998; Figure 1) were got from the Protein Data Bank (Berman et al., 2000) in pdb format.All hetero atoms and water molecules are removed by utilizing PyMol (version 1.3) software package (DeLano, 2002), while the Swiss-PDB viewer program (version 4.1.0)was used for energy minimization of the protein (Guex & Peitsch, 1997).The drugs were then tested against 4HBT and 1A7G in a molecular docking study.PyRx software (version 0.8) (Dallakyan & Olson, 2015) was used to create molecular docking simulations using the protein as a macromolecule and the drug as a ligand.
To enter the protein and ligands for the docking study, AutoDockVina was utilized, and AutoDock Tools (ADT) of the MGL software package was used to convert pdb into a pdbqt format.In AutoDockVina, the grid box was preserved at 39.1062, 52.4258, and 49.8323 for (4HBT) and 26.3257, 37.1107, and 36.5116 in the X, Y, and Z directions, respectively.After finishing docking, both the macromolecule and ligand structures were saved in the pdbqt format, which Accelrys Discovery Studio (version 4.1) used to examine and visualize the docking result, as well as search for non-bonding interactions between ligands and amino acid residues of receptor protein (Version ADS, 2017).The Ramachandran plot (Figure 1) demonstrated that 93.4% (4HBT) and 96.0%(1A7G) of residues were in the authorized region and no residues were missing when using the PDBsum online server to examine the validation of the proteins 4HBT and 1A7G.

Docking validation protocol
By removing the protein's co-crystallized ligand (PDB ID: 4HBT) and re-docking it into the same place, the docking results were confirmed.The co-crystallized ligands and the lowest energy posture achieved upon re-docking were superimposed, and their root mean square deviation (RMSD) was calculated.The RMSD must be within the trusted range of 2 Å for the docking operation to be validated (Onodera et al., 2007;Warren et al., 2006).To evaluate the precision of the calculated poses, the root-mean-square deviations (RMSDs) between the experimental and computational ligand structures were computed.
Calculated poses with heavy-atom RMSDs of less than or equal to 2.0 Å were regarded as reliable poses, as was the experimental structure that was retrieved from the PDB.We evaluated the success of docking and compared whether a reasonable pose was obtained or not because it was difficult to establish a proper match with any of the docking programs and scoring methods.

Molecular dynamics simulations
Molecular dynamics simulation investigations were undertaken using the Desmond module of Schrodinger to examine the vibrant binding behavior and binding consistency of protein-inhibitor complexes docked in various poses along with the apo protein (without ligand) (Bowers, 2006;Research et al., 2021).The complexes were first added to SPC water box, extending 10 Å beyond any of the complex's atoms, which were neutralized with Na/Cl À ions.To lower high energies in the projected model, the potential energy of the protein-ligand complexes was further improved and integrated into the NPT ensemble.The OPLS4 force field was used to execute the molecular simulation at 300 K temperature and 1 atm pressure over a 100 ns NPT production run.The simulation was extended further to 200 ns for all the complexes to analyze the behavior of all the complexes in long run.Finally, the stability of the protein-ligand complex was evaluated utilizing the ''Root Mean Square Deviation (RMSD)'' method, and structural fluctuation were determined using the ''Root Mean Square Fluctuation (RMSF)'' method (Krieger et al., 2004;Krieger & Vriend, 2015).

Pharmacokinetic and bioactivity analysis
The admetSAR 2.0 (http://lmmd.ecust.edu.cn/admetsar2/about)server was utilized to investigate the pharmacokinetic characteristics and toxicity of MGP esters.The pharmacokinetics parameters relevant to drug absorption, metabolism, and toxicity of the parent drug and its developed analogs were evaluated using the admetSAR online database (Cheng et al., 2012).AdmetSAR predicts the most recent and comprehensive hand-curated data for a variety of compounds connected with known ADME/T profiles using structure similarity search methods.Lipinski's (rule of five) is usually used to determine drug-likeness (Lipinski et al., 1997).To evaluate the ''drug-like'' properties of lead compounds, the Molinspiration online server (https://www.molinspiration.com/cgi-bin/properties)was used.The Molinspiration cheminformatics engine enables quick biological activity prediction.To screen for active chemicals, it is necessary to seek fragments or substructure characteristics.

Binding free energy Calculation using MM-GBSA
To quantify the binding affinity between the protein and small molecules, the molecular mechanics-generalized born surface area (MM-GBSA) approach was adopted.The prime module of Schr€ odinger was used to estimate the theoretical binding free energies of the promising NS2B-NS3 binding small molecules.The hits obtained after the XP docking, based on the glide scores, were subjected to MM-GBSA analysis.

Results and discussion
Eleven MGP esters have been changed in the current investigation using various aliphatic and aromatic chains (2-10) (Figure 2; Alen & Schols, 2012) and were attempted to perform anti-b-lactamase and antiviral assessment and geometrical optimization to realize the mode of their antimicrobial behavior.Following that, the observed behaviors were explained by performing thermodynamic, partial atomic charge, molecular docking, and combination in silico pharmacokinetic and drug-likeness calculations on each activity.Computer-Aided Drug Design (CADD) opened as a unique means of discovering emerging lead compounds and for accelerating the developments of possible drugs for a wide range of diseases.Nowadays, a wide range of computational investigations are being used to identify interaction between a molecule and a surface.Applications of CADD lead to drug discovery are developing on research basis.The current interest in drug design is to rationally discover promising therapeutics with multi-targeting effects, notable efficacies, and minimum side effects, especially in terms of toxicity.

Structural designation of designed MGP esters
Table S1, Figures S1, and 2 show the atomic identification and structural change of substituted MGP esters.Different aliphatic chains (C4-C16) and aromatic groups (trityl, cinnamoyl, and 4-nitrobenzoyl) were added to the hydroxyl (-OH) group of tor to detect the variance in biological activity.

Antiviral activity prediction
We choose to anticipate the antiviral activity of these carbohydrate esters 2-10 and compare them with Azidothymidine (AZT, an antiviral medication) using an antiviral application to determine the antiviral mode of inserted aliphatic and aromatic groups (Table 1).The modified MGP esters (2-10) exhibit potential antiviral efficacy when compared to their parent molecule, according to the projected antiviral activities.Comparing the potential scores of the aliphatic esters (2-7) and aromatic esters (8-10) to those of common medications like azidothymidine (AZT).

Physicochemical analysis
Thermal and molecular orbital properties are among the structural characteristics that are strongly influenced by a straightforward change in the chemical structure.The values of free energy and enthalpy can be used to calculate the spontaneity of a reaction and the stability of a result (Cohen & Benson, 1993;Ghasedi et al., 2021).Highly negative values have a higher chance of becoming thermally stable.The dipole moment also affects the development of hydrogen bonds and non-bonded interactions in drug design.According to Table 2, all of the MGP esters in the current investigation had bigger negative values for all calculated attributes than the parent MGP, indicating that the ester group's attachment might enhance these molecules' interactions and binding with various microbial enzymes.

Atomic partial charge analysis
The partial atomic charge is important in molecular computations for things like a simplified depiction of a molecule's global charge distribution and forecasting its conformational behavior.To determine the charge distribution and intrinsic properties of the interactions in the designed structure, Mulliken charges analysis is performed.The structure and reactivity of a molecule are frequently influenced by the polarity of chemical bonds (Heinz & Suter, 2004).The polarity of the molecule is not precisely defined by the molecular dipole moment, which is a vector.Different ways for giving partial charges to interacting atoms inside a molecule have been presented.The partial charges of all drug interacting atoms were computed using Mulliken Population Analysis and NBO methods (Gross et al., 2002).They are the most often used methods of population analysis.Atomic charges are connected to dipole moments and molecule polarizability (Mulliken, 1955).In both procedures, all hydrogen atoms had a positive charge, while another electronegative element (O) had a negative charge, as expected (Figure 3).In ester 6 (C-2, C-4, C-17, and C-25) showed a greater positive charge found due to the presence of higher electronegative element oxygen (O-10, O-21, O-22, and O-40), and H-13 exhibited a higher positive value than the other hydrogen because of the oxygen atom of -OH group.Interestingly, esters 7 (C-2, C-4, and C-25), 7 (C-1, C-16, C-25), and 8 (C-1, C-16) displayed a positive charge in both methods due to the presence of the oxygen atom of the carbonyl group.Due to the presence of more electronegative nitrogen atoms than any other esters, all esters (C-16 and C-25) displayed the highest positive charge (O, N, and Cl).Esters (2-7) had aliphatic chains, and esters 6 and 7 have the longest carbon chains, hence it was discovered that these two esters performed better.Last but not least, we depicted esters 6 and 7 from both the aliphatic and aromatic series (Figure 3) since they had a distinct graphical representation compared to the other esters, which had a scribbled look.

Docking validation study
The determination of the protein-bound ligand conformation and re-docking of the co-crystallized ligand was used to validate the correctness of the docking process to assess the capability of docking algorithms.The overlaid view between the docked ligand conformation and the co-crystallized ligand conformation is well shown in Figure 4, and the RMSD is 3.235 � Å.When compared to the ones mentioned in the current study, it was discovered that the complex interacted with identical amino acid residues.The RMSD rises to a very high level when the large symmetric molecules in the binding site are exchanged during docking.Unlike the tiny molecules, which can quickly acquire low RMSD even when  distributed randomly.According to certain published studies (Jones et al., 1997;Kontoyianni et al., 2004;Kroemer et al., 2004), a new standard for the excellence of docking poses based on visual examination has been proposed.Figure 4 displays the 2D depiction of interactions between a derived docking pose and the experimental ligand conformation for visual assessment.This visual examination reveals interactions that are identical to those in the experimental binding mode, as seen in Figure 4.The obtained results showed that the use of visual inspection as a new reference is crucial and that the alone is not a trustworthy metric for the quality of docking poses for docking validation.This helped to demonstrate the docking protocol's validity and accuracy in part.

Molecular docking simulation
A sort of bioinformatics modeling known as molecular docking simulation deals with the interaction of two or more molecules to produce a stable adduct.The chosen proteins were found to interact with important amino acids thanks to the outstanding docking performance of the MGP esters (2-3, 7-10) in Table 3 and Table S2.The drug molecules were seen to bind within the active site of both macromolecular structures in the docked views (Figure 5).Comparatively, the aromatic esters outperformed the aliphatic esters in terms of binding score.Figures S2-S4 and Figure 6 show the 2D schematics, which were produced by importing docking results into the Discovery Studio Visualizer and lig-plot, which depict the interactions between the inhibitor and bordering residues of 4HBT and 1A7G.These diagrams demonstrate how the amino acids participate in the pattern of interactions between the ligand and protein with a significant contribution to the overall interaction energy.The majority of these interactions, which help to understand molecular recognition, involve hydrophobic contacts, Van der Waals interactions, hydrogen bonds,  electrostatic interactions, and interactions between a single atom and an aromatic ring.For these esters, the most noticeable hydrogen-bonded binding sites were Ser237, Ser130, Ser130, Glu166, Asn104, Asn110, His237, Ser323, and Asn130 which are the most common active site of 4HBT and 1A7G.Besides, Moreover, hydrophobic interactions pi-alkyl, pisigma, amide pi stacked, and pi-pi stacked were found for the residues Tyr105, Pro167, Trp326, and Gly238.It is clear from the displayed data; that the binding sites were almost similar for aliphatic esters in comparison with the log plot (Figure S1) where Asn104, Asn130, and His237 revealed the closest distances of 1.909 Å, 1.934 Å, and 1.770 Å.But they exhibited bonds with the crucial binding site with some new interactions such as alkyl, pi-sigma, p-p stacked, and p-p-T shaped.
Along with Trp326, all the esters displayed the maximum p-p interactions with the Trp328, Tyr105, And Pro268 denoting the tight binding with the active site.Reports suggest  that Trp326, Tyr105, And Pro268 residues are considered the principal component of the p-p stacked, and p-p-T shaped is responsible for the accessibility of small molecules to the active site.
It should be emphasized that the examined MGP esters bind into the active site of the target proteins, much like the common medication azidothymidine (AZT), which is required to stop the researched infections' proteins from mutating.Binding energy and binding mode were improved in esters (2-3 and 9-10) due to substantial hydrogen bonding.In MGP, it was shown that changes to the -OH group improved the p-p interactions with the amino acid chain on the binding site, whereas changes to their polarity produced hydrogen bond interactions.
The ester (9) of both proteins, which formed with the same residues Asn136, Lys137, Pro145, Glu166, Gln321, Thr293, and Ser323 was where the highest significant H-bonds were found.Hydrogen-bonds play a vital role in shaping the specificity of ligand-receptor interaction, drug design, molecular recognition, and biological activity (Perlstein, 2001).The surface of hydrophobic and Hydrogen bonds of ester ( 10) is consequently represented in Figure 7.

Molecular dynamics simulations (MD)
To determine the dynamic stability of the protein-ligand complexes using their docked postures, MD simulations were carried out.On the top hits that were chosen and had high binding energies, MD simulations were run.The predicted conformational changes from the original structure across the simulation period were shown as root mean square deviation (RMSD).Additionally, root mean square fluctuation (RMSF) measurements were used to indicate structural stability, atomic mobility, and residue flexibility during the interaction of a protein-hit.The RMSD values for Compounds 4, 6, and 10 were calculated and given in Figure 8a-c, respectively.RMSD of Compound 10 showed a small deviation initially at almost 16 ns and then there was no significant fluctuation throughout the simulation and the backbone was consistent and in extended simulation for more than 100 ns, there was a little deviation of almost 1 Å from 2 to 3 Å and after 150 ns (extended) the simulation converged (Figure 8g).Similarly, estimated RMSF values less than 3 Å indicated high stability of the complex.Changes in conformation may be the cause of the deviation.
The residues from HIS 327 and LYS 336 showed fluctuation (Figure 8f).The RMSD of compound 6 showed variation initially till almost 50 ns and then there was not much abnormal deviation observed throughout 100 ns and then in extended simulation for 100 ns more, there was a flip at almost 183 ns (extended) and all other deviations were in range (Figure 8h).The deviation may be due to the flexibility of the ligand.For RMSF there was a fluctuation of almost 2.5 Å from HIS 327 and LYS 336 from HIS 327 and LYS 336 and the remaining structure was stabled comparatively and made contacts with ligand (Figure 8e).Compound 4 showed a very small deviation at almost 20 ns and it was consistent and conserved over the molecular dynamics simulation even in extended simulation from 100 to 200 ns there was no abnormal deviation observed throughout the simulation and was equilibrated till 200 ns (Figure 8i).The variation of RMSF was observed the same as for compound 6 and compound 10.The fluctuation was not abnormal where the ligand made contacts or hydrogen bonds (green lines) (Figure 8d).RMSD of apo-protein (without ligand) showed a small deviation from 1.2 to 2.2 Å between 50 ns to almost 75 ns and then there was no significant deviation throughout the simulation and the backbone was consistent and system was equilibrated till 100 ns (Figure 9a).Almost the same type of deviation was observed in protein when it attached with ligand, likewise the fluctuation in RMSF showed similar behavior as observed in protein-ligand complexes (Figure 9b).

Pharmacokinetic and drug-likeness properties
Certain fundamental characteristics of molecules, such as partition coefficient (logP), molecular weight (MW), and several hydrogen bond acceptors/donors are connected to the Lipinski 'rule of five'.The SwissADME program was used to calculate drug-likeness predictions, which are listed in Table 4 as a whole.
In comparison to high molecular weight derivatives, low molecular weight derivatives are easily absorbed, dispersed, and transported.Except for a few exceptions, as molecular weight rises, the bulkiness of the molecules rises in tandem.The TPSA (Topological Polar Surface Area) is a critical physicochemical feature of a molecule that indicates the polarity of a substance.This parameter was checked to see if it might be used to analyze drug transport qualities.The sum of all polar atoms, primarily oxygen and nitrogen, as well as connected hydrogen, is known as the polar surface area.A mixture of GPCR, ion channel modulators, kinase inhibitors, nuclear  receptor ligands, protease inhibitors, and enzyme inhibitors has been used to predict the efficacy of compounds to qualify for therapeutic development (Table 5).The larger the bioactivity score, the higher the probability of the specific molecule being active.A number between 0.50 and 0.00 is regarded as moderately active, while a score of less than À 0.50 is assumed to be inactive.If a molecule's bioactivity score is more than 0.00, it has promising biological activities.
The bioactivity score provides details on the drug binding cascade that is used to build a new functional medicine with a higher binding selectivity profile and fewer negative side effects.The toxicity profiles of all chosen MGP esters were assessed, and they are listed in Table 6.

Binding free energy calculation using MM-GBSA
MM-GBSA approach was adopted to calculate the binding free energies of compounds 4, 6, and 8 to assess their binding affinity with receptor protein.The calculated binding free energies DGbindfor the compounds 4, 6 and 8 are À 60.51, À 53.95 and À 78.06 kcal/mol, respectively.

Discussion
The study of the nine previously designed antimicrobial MGP esters 2-10 was conducted through in silico computational approach.In this view, the binding scores and stability of the protein-ligand complexes of MGP esters with the CTX-M-15 b-lactamase and HPV 31 were identified along with predicting their antiviral, physicochemical, bioactivity, and pharmacokinetic properties.Geometry optimization of the MGP esters was performed through density functional theory (DFT) at B3LYP/6-311 þ G (d,p) level of theory (Becke, 1988) to predict their thermochemical and physicochemical parameters such as electronic energies, enthalpies, entropies, polarizability, natural bond orbital (atomic partial charge), etc. (Heinz & Suter, 2004).Among all the designed esters, ester 7 displayed the highest value for polarizability (710.389a.u.), heat capacity (386.130cal/mol-kelvin), and entropy (461.027cal/mol-kelvin), where the aromatic ester 9 revealed the highest energy score (9964.125Hartree).The partial atomic charge analysis revealed that a notably reduced electron density characterizes the single C-C bond of the substituents, which results in a decreased interaction strength.In the field of partial atomic charge analysis, the greater the electronegativity value, (i) the more intensive the interaction between electron donors and electron acceptors, and (ii) the greater the extent of conjugation of the whole system (Cao et al., 2016;Hosen et al., 2021).All the aromatic ringsubstituted derivatives displayed a comparatively higher score than the aliphatic one.Finally, it was proved that modification of hydroxyl (-OH) groups of MGP significantly increased its thermodynamic properties, which indicated the inherent stability of the designed derivatives (Maowa et al., 2021;Farhana et al., 2021).Moreover, the presence of bulky acylating groups also suggested a possible improvement of these physicochemical properties (Hosen et al., 2022;Kawsar et al., 2022).The antimicrobial mode of designed MGP esters was checked by the prediction of antiviral parameters and a comparison was made with azidothymidine (AZT, a standard antiviral drug) using an online-based antiviral predictor (Table 2; Farhana et al., 2021).Antiviral activity analysis exhibited that the entire series of modified MGP esters (2-10) had a promising antiviral potentiality; however, in special cases, these esters were more effective against the Hepatitis B virus (HBV) and Human herpesvirus (HHV).Interestingly, aromatic ring substituted esters (8-10) revealed a better activity score than the aliphatic series (2-7) in comparison with standard drugs AZT.According to these results, it is reasonable to propose that the substitution of an aromatic ring may increase the antimicrobial capability of a carbohydrate ester (Chowdhury et al., 2016;Kawsar et al., 2015;Rana et al., 2020).Keeping in view the features mentioned above, we propose herein to study the antimicrobial behavior of the nine designed esters through molecular docking simulation with b-lactamase of E. coli (4HBT) and human papillomavirus type 31 (1A7G).A total of 10 complexes for two proteins with eight MGP esters were docked and two types of proteins including b-lactamase of E. coli (4HBT) and human papillomavirus type 31 (1A7G) were selected.Both antibacterial activity and b-lactam antibiotic resistance are mediated by beta-lactamase, where, 1A7G is responsible for viral DNA genome replication, regulation of DNA replication, transcription, and DNA binding.The outcomes of the docking analysis showed that all esters, along with the parent compound had binding affinities ranging from À 4.2 to À 9.3 kcal/mol for b-lactamase and À 4.0 to À 7.8 kcal/mol for human papillomavirus type 31 (1A7G).As shown in Table 3 and Table S1, MGP esters (2-3 and 7-10) displayed comparatively improved and higher binding affinity than the parent MGP.These results indicated that insertion of a long carbon chain/aromatic ring molecule improved the binding affinity; while, it decreased as the chain grew longer.Due to having high electron density, the aromatic derivatives can strongly bind with the receptor protein and revealed promising binding energy.But there is an exception that compound 8 showed lowest binding energy and its maybe caused steric hindrance the of nine phenyl ring presence in the compounds.In addition, adding aromatic groups like tri-phenyl, 4-NO 2 -benzoyl and 4-Br-benzoyl made some standard improvements in binding affinities.However, modification with halogenated aromatic rings increased the binding affinity.Aliphatic chain substituted esters (2-3) exhibited a smooth rising of the binding score (À 7.7, À 6.2 À 6.5 and-4.9kcal/mol) for both the protein and interacted with the most common active site of the proteins.Still, the binding score was notably decreased as aliphatic chain length increased for the esters (4-7) in the case of both proteins.This observation indicated that a small aliphatic chain may improve the binding mode as well as the biological efficacy of MGP esters.Aromatic esters (9-10) displayed a higher binding score (À 6.7 and À 9.3 kcal/mol for 4hbt and À 6.7 and À 7.8 kcal/mol for 1A7G) in comparison with aliphatic esters.Here is a clear indication that due to having high electron density, aromatic substituents can easily enhance the binding capability as well as the antimicrobial activity of the MGP esters in comparison with standard drugs AZT.The hydrogen bond is an important factor that can promote the binding specificity of the drug with the protein and all designed esters interact mostly by the formation of hydrogen bonds (Gross et al., 2002).It was found that the analyzed MGP esters were bound within the active site of the b-lactamase of E. coli (4HBT) and human papillomavirus type 31 (1A7G), which is essential to prevent the activity of these pathogens.The docking findings indicated that the esters (2, 3, 9, and 10) may be extremely important for human papillomavirus type 31 (1A7G) and potential antibacterial agents to overcome the challenge of the bacterial b-lactamase enzyme.Also, docking with 4HBT was validated by visual inspection.Molecular dynamics simulations were performed on the top hits containing high binding energies that were selected.The predicted conformational changes from the original structure across the simulation period were shown as root mean square deviation (RMSD).Additionally, root mean square fluctuation (RMSF) measurements were used to indicate structural stability, atomic mobility, and residue flexibility during the interaction of a protein-hit.The pharmacokinetic and bioactivity prediction clarified the druglike properties of the MGP esters.Table 4 exhibits esters (2-4) following the rule of five indicating the good bioavailability of MGP esters.It was found that all the esters have good hydrogen bonds donor and acceptor which is consistent with Lipinski's rule of five.The more the value of topological polar surface area (TPSA) the more the polarity of the compounds, which should be less than 140 Å 2 .The TPSA in these molecules was fitted (except ester 8) most of the druglikeness rules (Krieger & Vriend, 2015).The bioactivity score of all designed MGP esters is displayed in Table 5.The obtained values of bioactivity score exhibited that esters (2-4 and 9-10) followed the promising drug likeness.Furthermore, the prediction of toxicity for the MGP esters (Table 6) revealed that all compounds are non-toxic and non-carcinogenic with a level III of oral toxicity.In addition, no violations were found with these MGP esters using pan assay interference compounds (PAINS).Chemical substances known as PAINS frequently produce false-positive findings in high-throughput screens.PAINS tend to react nonspecifically with a variety of biological targets rather than affecting a single desired target.Our studied compounds showed promising pharmacokinetic parameters in comparison with standard drugs Azidothymidine (AZT).More drug-likeness in vitro and in vivo experiments, including ones examining nontoxic concentrations toward healthy cells, may soon be done to this intriguing inquiry.
The outcomes clarified that the insertion of several electron-enriched and electron-deficient groups firstly in the C-5 position and finally on the C-2, C-3, and C-4 positions, of MGP structures, markedly enhances the antimicrobial ability of the designed compounds.The external membrane of the gram-negative organism consists of phospholipids, that have a pure hydrophobic property.All the Compounds were modified with a long hydrocarbon chain and aromatic ring at the C-5, C-2, C-3, and C-4 positions which made it capable of hydrophobic interaction.As a result, hypothetically it may suggest that MGP esters made hydrophobic interaction with the outer phospholipid membrane of bacteria.Advanced computational applications have been shown to be effective tools and notable successes have been achieved using these techniques.CADD is a specialized discipline, whereby different computational methods, for example, molecular docking and molecular dynamics are used to simulate interactions between receptors and drugs in order to determine binding affinities.

Conclusions
In the present study, nine MGP esters (2-10) were studied based on computer-aided drug design, including molecular optimization, antiviral activity prediction, molecular docking, pharmacokinetics, and drug-likeness determination.In this investigation, the inherent characteristic stability and biochemical behavior of MGP and its esters were studied to suggest potential antimicrobial agents.The most significant properties for biological chemistry and chemical reactivity studies like antiviral prediction, Mulliken charges analysis, and physicochemical properties were evaluated, which indicates that they may be good drug molecules.All of the designed MGP esters outperformed MGP in terms of physicochemical properties and were more reactive than the parent drug.The MGP structure can be greatly improved by adding different aliphatic and aromatic groups to change its biological activity mode.The MGP esters 2-10's antiviral predictions demonstrated their potential against different infections.These data were explained by molecular docking, which showed the MGP esters' promise antibacterial effectiveness.According to molecular docking analysis, esters 2, 3, 9, and 10 may be extremely important in their ineffectiveness against human papillomavirus type 31 (1A7G) and bacterial b-lactamase (4HBT).The protein-ligand complex is quite stable in a biological system, according to this study, which strongly and unambiguously supports molecular dynamics investigations up to 200 ns.This result also verifies the docked complex's binding stability in the trajectory analysis.It might be successfully revealed that the esters of 9 and 10 could show promising antimicrobial activity, and it could be said that this could be due to the presence of an electron-withdrawing aromatic ring, which could enable the formation of a strong H-bond with protein, even if other conventional bonds are also possible.In the case of large, nearly symmetrical molecules, RMSD is not a good parameter in the docking validation technique.Visual examination produced incredibly convincing results in this investigation's molecular docking validation approach.These esters pharmacokinetic and drug-likeness characteristics were examined, which demonstrated that the combination of toxicity, bioactivity, and drug-likeness had promising outcomes because the majority of the designed molecules have improved kinetic parameters and it maintains all drug-likeness rules as well as an intriguing result in terms of biological activity.

Figure 4 .
Figure 4. Superimpose view of re-docking RMSD value of 3.235 Å (Green ¼ Docked, blue ¼ Original) in the active site using PyMOL.

Figure 6 .
Figure 6.Non-bonding interactions of compounds (9 and 10) with the amino acid residues of 4HBT and 1A7G generated by Discovery Studio.

Figure 8 .
Figure 8.The MD simulation trajectories from 200-ns simulation time, blue colour represents the protein and red represent the ligand, here (a) RMSD plot of compound 4-complex; (b) RMSD plot of compound 6-complex; (c) RMSD plot of compound 10-complex; (d) RMSF plot of compound 4-complex; (e) RMSF plot of compound 6-complex and (f) RMSF plot of compound 10-complex.(g) The extended MD simulation trajectory of compound 10-complex from 100 to 200 ns simulation time.(h) The extended MD simulation trajectory of compound 6-complex from 100 to 200 ns simulation time.(i) The extended MD simulation trajectory of compound 4-complex from 100 to 200 ns simulation time.

Table 4 .
Prediction of drug-likeness properties of MGP esters.

Table 5 .
The drug-likeness score of MGP derivatives through molinspiration cheminformatics online server.