In silico studies of benzimidazole derivatives as sustainable inhibitors against Methicillin-resistant Staphylococcus aureus

Antimicrobial resistance is becoming more rampant in our world today, and different measures are being taken to combat this challenge. Benzimidazoles are classified as heterocyclic compounds with notable pharmacological properties. As a result, benzimidazole has been combined with other compounds that have remarkable actions to create a more potent molecule. Exploring these substances to combat antibacterial resistance would therefore aid in achieving good health and wellbeing and promote sustainable development. Predicting the effectiveness of the compounds before manufacturing and clinical testing has made drug design easy. This study employs in silico methods like molecular docking to investigate alternate antibacterial agents from a library of benzimidazole derivatives. A library of compounds with a benzimidazole template was screened against the three-dimensional (3D) structure of peptidoglycan transpeptidase (PPB2A) of Staphylococcus aureus. Two binding sites were identified in the protein: the main site and the allosteric site. Molecular docking was done on the main and allosteric sites to obtain free binding energy ranging from -7.3 to -5.8 and -4.9 to -4.5 kcal/mol, respectively. The predictive Absorption, Distribution, Metabolism, Excretion, and Toxicity (ADMET) experiments were done on the compounds to ascertain their safety. The results were compared to those of known antibiotics, and the compounds performed effectively. The benzimidazole derivative can be adopted as a prospective antibacterial agent with an alternative pathway for combating resistance issues and enhancing the quality of health and well-being globally.


Introduction
Antimicrobial resistance is one of the top 10 difficulties the public health sector faces, among many others.Due to the growing rate of resistance of microorganisms such as bacteria, fungi, parasites, and viruses to drugs and drug-like compounds, effective substitutes must be discovered [1][2][3].According to global estimates, antimicrobial resistance was responsible for approximately 2.4 million deaths in 2019 [4].Some of the common bacteria that cause some chronic diseases include Mycobacterium tuberculosis associated with tuberculosis [5], Escherichia coli associated with intestinal infections, Staphylococcus aureus, Pseudomonas aeruginosa, and Klebsiella pneumonia [6,7].S. aureus is one of the most common gram (+) bacteria, it is responsible for a wide range of clinical infections including skin and soft tissue infections that can develop into bone infections or severe muscle pain [8].It can also affect the lungs and heart valves if not treated.They are significant contributors to food-borne illness [8].
Studies show that the bacterial cell wall of S. aureus contains peptidoglycans which are the main building blocks and thus influence its survival during cell division and growth [8,9].In S. aureus, resistance is initiated by the Penicillin-binding protein (PBP2A), a peptidoglycan transpeptidase in conjunction with the PBP2 transglycosylase domain when β-lactam antibiotics are introduced [9].As a result, researchers are looking for compounds that can inhibit the activity of these enzymes.Organic compounds, particularly heterocycles, have attracted attention due to their powerful pharmacological properties [10].This group includes benzimidazole derivatives, which are composed of an imidazole fused with a benzene ring [11,12].They have found numerous applications as antimicrobials [13], anti-cancer [14], anti-tubercular [15], anti-oxidant [16], antimalarial [17], etc.Some commercially available benzimidazole derivatives used in the treatment of various diseases are shown in Fig. 1.In order to shorten the 12 to 15-year process of drug design and development, Computer-aided Drug Design (CADD) has offered sustainable solutions to overcome the hurdles involved in the initial stage.There are several tools available for designing compounds, such as in silico ADMET predictors, molecular docking, and molecular dynamic simulation, which allows us to test the compounds on proteins responsible for an activity in the biological system [18].This study seeks to investigate alternative drug-like substances from benzimidazole derivatives against the 3D structure of peptidoglycan transpeptidase (PBP2A) in Staphylococcus aureus as a sustainable treatment option for multidrug resistant strains.

Protein preparation
The experimentally validated 3D structure of PBP2A of Methicillin-resistant Staphylococcus aureus (MRSA) was downloaded from the Protein data bank (PDB) using the PDB ID: 4CJN as reported by [19].The protein was then prepared using the chimera software by removing non-amino residues, minimizing the structure, and adding Gasteiger charges.The protein was taken further for the docking and simulation studies.

Ligand preparation
The benzimidazole template was utilized for the compound search on the PubChem database [20].The compounds were screened using Lipinski's rule of five (LO5), and a total of about 1510 compounds were downloaded alongside four (4) recognized antibiotics in Structure-Data Files (SDF) format.The Open Babel in PyRx was then used to convert these compounds into Autodock docking formats (pdbqt).

Virtual screening and post-docking analysis
The prepared ligands and known antibiotics (gentamicin, penicillin G, streptomycin, ampicillin) were screened against PBP2A using Autodock vina [21].The active site for the main and allosteric sites was set using the amino acid residues identified by [22].Using Discovery Studio, the binding interactions between the top five hits, and four antibiotics were depicted.

In silico toxicity prediction and drug-likeness
The in silico toxicity and drug-likeness were analyzed using OSIRIS Property Explorer and ADMETlab webserver [23].The prediction took Lipinski's rule of five (LO5) into account when assessing the pharmacokinetic and toxicological profiling of the compounds and known antibiotics [24].

3D Structure of Protein
The protein has two chains with three domains.The domains are the PBP2A domain (containing 2-114 amino acids sequence), N-terminal transpeptidase domain (122-283 amino acid sequence), and transpeptidase domain (320-631 amino acids sequence) (Fig. 2).The transpeptidase domain containing the Ser403 amino acid responsible for resistance in MRSA was selected as the main active site.The protein had an allosteric site which serves as a regulatory system that affects the activity at the main binding site [8,9,25].

Virtual screening and post-docking analysis
A total of 1510 compounds were docked against the main site and allosteric site of PBP2A.The 5 top hits compounds in the main site were reported to have binding energy ranging from -7.3 to -7.1 kcal/mol whereas that of the allosteric site ranged from -5.4 to 5.0 kcal/mol (Table 1).
With the help of BIOVIA Discovery studio, the 2D and 3D interaction between the amino acid residues of PBP2A and the atoms of ligands was observed (Fig. 3).The ligand with the highest binding affinity, 1023408, interacted with Ser 461, Ser 462, Asn 464, Thr 600, and Glu 602.
The NH linker carrying the imidazole and methylene groups that were joined to the benzimidazole template helped to make the ligand more active.Similarly, for ligand 3079203, the contact between atoms and Tyr 529, Thr 444, and Glu 602 was facilitated by the NH present in the benzimidazole molecule and the azanecarboximidamide fragment.Ligand 137054718 formed a hydrogen bond with the NH on the benzimidazole using Tyr 441 while at the allosteric site, the ligand 137054718 was also the third best.The ligands functioned comparably to the well-known antibiotics, despite streptomycin and ligand 137054718 having identical binding affinities whereas, at the allosteric site, all the ligands performed better than the known especially Pencillin G which is a known β-lactam antibiotic.Penicillin G -6.9 Penicillin G -4.9 Ampicillin -6.5 Ampicillin -4.9 Gentamicin -5.8 Gentamicin -4.5

In silico toxicity prediction and drug-likeness
According to OSIRIS Property Explorer's predictions for in silico toxicity and drug-likeness, all the compounds have a variety of drug-relevant properties that fall within acceptable ranges (molecular weight [MW] ⩽ 500 g/mol; TPSA ⩽ 160 Å; clogP ⩽ 5; logS > -4 mol/dm3) as shown in Table 2.The molecular weight depicts how heavy a molecule is, streptomycin went over the allowable limit of 500 g/mol.When considering drug bioavailability, the topological polar surface area is crucial.the value exceeds 160 Å, it will have an impact on the hydrogen bonding of the compounds.All the screened compounds fall within this range of the partition coefficient (clogP), which indicates that they have good absorption properties.Similarly, the solubility prediction is associated with absorption and distribution properties and all the screened compounds within the range.More specifically, toxicity characteristics such as impact on the reproductive system, irritability, and tumorigenicity are assessed.The top 5 hits of benzimidazoles didn't pose any toxicity danger, the structures are shown in Fig. 4. Penicillin G is the only mutagenic or tumorigenic substance that exhibits high risk.None of the substances, except for streptomycin, demonstrated a high risk to the reproductive system.Toxicology and other physicochemical properties add up to create the drug score.A ligand becomes a possible drug candidate if its drug score is greater.

Figure 4 :
Figure 4: Chemical Structures of the Best Five (5) Hits with their PubChem ID4.Conclusion and Recommendation

Table 1 :
Docking score of top 5 best hits and known antibiotics in PBP2A.

Table 2 :
Physicochemical properties and toxicity risks of top 5 hits in comparison to known antibiotics.