Fighting Antibiotic Resistance: New Pyrimidine-Clubbed Benzimidazole Derivatives as Potential DHFR Inhibitors

The present work describes the design and development of seventeen pyrimidine-clubbed benzimidazole derivatives as potential dihydrofolate reductase (DHFR) inhibitors. These compounds were filtered by using ADMET, drug-likeness characteristics calculations, and molecular docking experiments. Compounds 27, 29, 30, 33, 37, 38, and 41 were chosen for the synthesis based on the results of the in silico screening. Each of the synthesized compounds was tested for its in vitro antibacterial and antifungal activities using a variety of strains. All the compounds showed antibacterial properties against Gram-positive bacteria (Staphylococcus aureus and Staphylococcus pyogenes) as well as Gram-negative bacteria (Escherichia coli and Pseudomonas aeruginosa). Most of the compounds either had a higher potency than chloramphenicol or an equivalent potency to ciprofloxacin. Compounds 29 and 33 were effective against all the bacterial and fungal strains. Finally, the 1,2,3,4-tetrahydropyrimidine-2-thiol derivatives with a 6-chloro-2-(chloromethyl)-1H-benzo[d]imidazole moiety are potent enough to be considered a promising lead for the discovery of an effective antibacterial agent.


Introduction
The treatment of nosocomial infections poses a significant global risk to public health due to drug-resistant bacteria, such as methicillin-resistant Staphylococcus aureus (MRSA) and multidrug-resistant Escherichia coli [1][2][3]. Research commissioned by the United Kingdom Government estimates that "the cost in terms of lost global production between now and 2050 would be an astounding 100 trillion USD" if no action is taken. Infections caused by fungi may significantly threaten human health, especially for immunocompromised patients. When it comes to clinical care, invasive fungal infections (IFIs) pose a significant challenge on a global scale [4][5][6]. It is imperative that attempts to discover new antibiotic agents be stepped up to keep pace with the worrisome increase in cases of antibiotic resistance being demonstrated by disease-causing microbes [7,8]. One crucial step is the identification of potent inhibitors of receptors that are critical to the bacteria's survival.  The absorption parameters of the molecules are illustrated in Table 3. As a model of how medications are absorbed by the human digestive tract, the human colon epithelial cancer cell line, known as Caco-2, is employed. This model is useful for determining whether a substance is appropriate for oral administration, predicting intestinal permeability, and researching drug efflux. Caco-2 permeability is optimum when the value is higher than −5.15 log units, and fortunately, all the molecules displayed optimum Caco-2 permeability [25]. It is possible to acquire a better knowledge of the process of drug efflux with the aid of MDCK-MDR1 cells, which also draws attention to early potential problems with drug permeability. It has been discovered that the permeability of MDCK-MDR1 may, in addition to intestinal permeability, be used as an accurate predictor of the permeability of the bloodbrain barrier [26]. Many of the molecules displayed P-gp-inhibitor and P-gp-substrate activity. All the designed molecules displayed excellent human intestinal absorption (HIA). The molecules' bioavailability of 20% and 30% were within acceptable limits. Table 3. An absorption parameter of methyl 2-((6-chloro-1H-benzo[d]imidazol-2-yl)methylthio)-1,2,3,4-tetrahydro-6-methylpyrimidine-5-carboxylate derivatives (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35)(36)(37)(38)(39)(40)(41)  The distribution and metabolism profile of the molecules are depicted in Table 4. The plasma protein binding (PPB, <90%) drugs with high protein bound within them may have a low therapeutic index. Many of the molecules displayed a PPB at less than 90%. The volume distribution (VD; optimal 0.04-20 L/kg) of all molecules was within the acceptable limit range. None of the molecules displayed a BBB penetration potential. Cytochrome (CYP) enzymes play an important role in drug metabolism; therefore, their substrate or inhibitor contributes to the drug's action. None of the molecules in the current study demonstrated CYP inhibitory or substrate potential [27]. An excretion and toxicity profile of the molecules is tabulated in Table 5. Many of the molecules displayed a moderate to low clearance (CL, High: >15 mL/min/kg; moderate: 5-15 mL/min/kg; low: <5 mL/min/kg) rate. All the molecules exhibited a short half-life (T 1/2 , <3 h). The toxicity profile of the molecules suggested favorable properties, and many of the values were within the range. The physicochemical radar of the developed molecules obtained from the ADMETlab 2.0 web server is reported in Figure 2, which indicates the molecules' favorable physicochemical parameters to be developed further [27]. Most of the developed molecules displayed physicochemical properties within the upper limit of the acceptable range, as per the radar images. The physicochemical radar contains almost all the properties that are ideal for the development of any lead as a potential therapeutic agent. An environmental toxicity profile (bioconcentration factors, IGC 50 , LC 50 FM, and LC 50 DM) of the designed molecules is shown in Table 6. The environmental toxicity profile of the molecules was optimum and within the acceptable range.

Molecular Docking
In the docking calculations, comparisons have been made between the binding affinities of the designed derivatives and the binding mode of the native ligand that is found in the crystal structure of DHFR (PDB ID: 5CCC). The molecular interactions of the titled compounds are exemplified in the Supplementary Information; in Table 7, the most potent compounds' 2D-and 3D-docking poses are described. The native ligand displayed a binding affinity with DHFR of −8.5 kcal/mol, and it established six conventional hydrogen bonds with Asp27, Ala6, Ile5, and Arg57, in addition to one carbon-hydrogen bond with Ile94. It has established many hydrophobic interactions, such as Pi-sigma bonds, Pi-Pi T-shaped bonds, alkyl bonds, and Pi-alkyl bonds with Ile50, Phe31, Ile94, Ile5, and Ala7. Compound 27 exhibited a binding affinity value of -8.6 kcal/mol with the formation of five hydrophobic bonds (pi-sigma, alkyl, pi-alkyl) with Leu28, Lys32, Leu28, Ala7, and Phe31. Compound 29 exhibited a binding affinity value of −9.3 kcal/mol with the formation of one hydrogen bond and several hydrophobic bonds (pi-sigma, pi-pi T-shaped alkyl, pi-alkyl) with Leu28, Phe31, Ile50, Ile5, Ala7, Met20, Trp30, and Phe31. Compound 30 displayed a binding affinity value of −9.6 kcal/mol with the formation of one carbon-hydrogen bond and many hydrophobic bonds (pi-sigma, pi-pi T-shaped alkyl, pi-alkyl) with Leu28, Phe31, Ile50, Ile5, Ala7, Met20, Trp30, and Phe31. Table 7. The 2D-and 3D-docking postures of molecules selected for the synthesis.

41
Compound 33 displayed a binding affinity value of −9.0 kcal/mol with the formation of two conventional hydrogen bonds and one carbon-hydrogen bond with Met20, Ile94, and Asp27. It also displayed many hydrophobic bonds (pi-sigma, pi-pi T-shaped alkyl, pi-alkyl) with Leu28, Leu54 Phe31, Ile50, Ile5, Ala7, Met20, Trp30, and Phe31. Compound 37 displayed a binding affinity value of −8.7 kcal/mol with the formation of one conventional hydrogen bond and one carbon-hydrogen bond with Ala7 and Asp27. It also displayed many hydrophobic bonds (pi-pi T-shaped, alkyl, pi-alkyl) with Tyr100, Ile50, Met20, Leu28, and Ile14. Compound 38 displayed a binding affinity value of −9 kcal/mol with the formation of one conventional hydrogen bond with Ile94. It also displayed many hydrophobic bonds (pi-pi T-shaped, amide-Pi stacked, alkyl, pi-alkyl) with Leu28, Met20, Ile5, Phe31, Ala7, and Phe31. It also displayed electrostatic interactions with Glu17. Compound 41 displayed a binding affinity value of -9 kcal/mol with the formation of one conventional hydrogen bond and one carbon bond with Thr113 and Trp30. It also displayed many hydrophobic bonds (pi-sigma, alkyl, pi-alkyl) with Leu28, Lys32, Ile5, Phe31, Ala7, and Phe31. Therefore, from the above results, the compounds that showed a binding affinity value lower than the native compound in the X-ray complex (27, 29, 30, 33, 37, 38, and 41) were selected for the synthesis and biological evaluation.

Synthesis of the Selected Compounds
Compounds 27, 29, 30, 33, 37, 38, and 41 were chosen for the synthesis based on the results of the in silico screening and molecular docking investigations.

In Vitro Antibacterial Activity
The findings of the synthetic derivatives' antibacterial and antifungal activities are listed in Table 8, which display the MICs and MFCs, respectively (n = 3).

In Vitro Antibacterial Activity
The findings of the synthetic derivatives' antibacterial and antifungal activities are listed in Table 8, which display the MICs and MFCs, respectively (n = 3).
Xue-Qian Bai et al. reported some pyrimidine derivatives as potential antimicrobial agents, where one compound presented the most potent inhibitory activities against Staphylococcus aureus, Escherichia coli, and Candida albicans, with a MIC of 2.4 µmol/L. Additionally, it was the most potent, with MICs of 2.4 or 4.8 µmol/L against four multidrug-resistant, Gram-positive bacterial strains [29]. Omaima G. Shaaban et al. synthesized and evaluated some 3,4-dihydrothieno[2,3-d]pyrimidine derivatives as potential antimicrobial agents. Many of the derivatives displayed half of the potency of levofloxacin against Pseudomonas aeruginosa and Proteus vulgaris and also half the activity of ampicillin against the Gram-positive bacterium B. subtilis [30]. In the current study, each of the selected and synthetized compounds was tested for its in vitro antibacterial and antifungal activities using a variety of strains. All the compounds produced showed antibacterial activities against Gram-positive bacteria (Staphylococcus aureus and Staphylococcus pyogenes) as well as Gram-negative bacteria (Escherichia coli and Pseudomonas aeruginosa). All the compounds had actions against Gram-positive and Gram-negative bacteria that were much more powerful than that of ampicillin. Most of the compounds either had a higher potency than chloramphenicol or an equivalent potency to ciprofloxacin. Against Escherichia coli, 33 and 41 were sensitive at 25 µg/mL, whereas 29, 37, and 38 were sensitive at 50 µg/mL. It was observed that 27 and 30 were non-sensitive against Escherichia coli. Infections brought on by the opportunistic bacteria Pseudomonas aeruginosa are often treated with the drug ciprofloxacin. In spite of the widespread administration of ciprofloxacin, the number of Pseudomonas aeruginosa strains that have developed resistance to the drug continue to rise [31]. Infections caused by Pseudomonas aeruginosa are notoriously difficult to treat because of the bacteria's high levels of inherent and acquired antibiotic resistance. Once Pseudomonas aeruginosa has taken hold in a human host, it quickly creates genetic changes that make it resistant to antibiotics and better able to adapt to the host environment [32]. Therefore, it is not surprising that ciprofloxacin displayed low sensitivity (25 µg/mL) against Pseudomonas aeruginosa. However, all the compounds were sensitive at 50 µg/mL except 38, which was non-sensitive against Pseudomonas aeruginosa. Against Staphylococcus aureus, 27, 33, 37, 38, and 41 were sensitive at 25 µg/mL, whereas 29 and 30 were sensitive at 50 µg/mL. Against Staphylococcus pyogenes, 30 and 33 were sensitive at 25 µg/mL, whereas 29, 37, 38, and 41 were sensitive at 50 µg/mL. Compound 27 was non-sensitive against Staphylococcus pyogenes.
Candida, a yeast, is developing increased resistance to antifungal medications. Candida infections may be challenging to treat because of the possibility of drug resistance. Two patients with oral candidiasis, who did not improve while using nystatin in conjunction with triamcinolone acetonide, are presented. High in vitro resistance to nystatin was seen when triamcinolone acetonide was used in conjunction with the Candida albicans isolates collected from the patients after therapy [33]. This might be a reason why nystatin demonstrated low sensitivity (100 µg/mL) against Candida albicans, whereas all the compounds were sensitive at 100 µg/mL, which is equipotent to nystatin and more potent than griseofulvin except for 33, which was sensitive at 200 µg/mL. Against Aspergillus niger, all the compounds were equipotent with nystatin and griseofulvin (100 µg/mL) except for 30, which was sensitive at 200 µg/mL and 37 was non-sensitive. Against Aspergillus clavatus, all the compounds were sensitive at 100 µg/mL except for 38 and 41, which were non-sensitive. It was observed that 29 and 33 were sensitive against all the bacterial and fungal strains.
From the above results, it was observed that these molecules have enough potential as antimicrobial agents. These molecules displayed an optimum binding affinity for the DHFR enzyme and showed significant inhibition. Therefore, we proposed that these molecules exert antimicrobial activity via the inhibition of DHFR.

Pre-ADMET Profile and Drug-Likeness Properties
The in silico ADMET assessment models are a new type of tool that has been created to provide medicinal chemists with extra support in the process of the creation and optimization of leads. ADMETlab 2.0 is a revamped version of the AMDETlab web server, which is commonly used for predicting the pharmacokinetics and toxic characteristics of various compounds (https://admetmesh.scbdd.com/ accessed on 21 August 2022) [27].

Molecular Docking
Using Autodock vina in PyRx 0.8, the hypothesized derivatives, as well as the native ligand, were docked to the crystal structure of the wild-type E. coli dihydrofolate reductase [34]. The structures of the proposed derivatives and the native ligand were drawn using ChemDraw Ultra 12.0. (Mol File format). By using the open-Babel tool, the ligands were imported into the PyRx software. By using the Universal Force Field (UFF), each of the ligands was optimized in terms of reducing the amount of energy [35]. The ligands were then converted to the PDBQT format and prepared for docking purposes. The crystal structure of wild-type E. coli dihydrofolate reductase was obtained from the RCSB Protein Data Bank (PDB) with PDB ID: 5CCC (https://www.rcsb.org/structure/5CCC, accessed on 28 August 2022). The enzyme structure was refined using Discovery Studio Visualizer (version 19.1.0.18287), and then it was purified and prepared for docking using the same program [36]. The output file of the enzyme was saved in a PDB file format and imported to PyRx to perform the molecular docking studies. In order to aid molecular docking, a three-dimensional grid box (size_x = 41.7862652138Å; size_y = 39.1754565902Å; size_z = 37.1398050256Å) with an exhaustiveness value of eight was developed [34]. The strategy reported in previous papers was used in order to carry out the complete molecular docking method as well as to locate cavities and active amino acid residues [18,28,[37][38][39][40][41][42][43][44][45][46]]. The exposed cavity of the DHFR is shown with the co-crystallized ligand molecule in Figure 4.

Chemistry
From the Lab Trading Laboratory in Aurangabad, Maharashtra, India, all the essential chemicals and reagents of synthetic quality were obtained. The progression of the reaction was monitored and verified using thin-layer chromatography (TLC, Merck precoated silica GF 254). The melting points were determined using a VEEGO Model VMP- were performed at concentrations of 200, 100, 50, 25, 12.5, and 6.250 µg/mL. A control that did not contain antibiotics was subcultured (before being inoculated) by distributing one loopful of media uniformly over a fourth of a plate of medium that was adequate for growing the test organisms. This was followed by overnight incubation at 37 • C. The lowest concentrations of derivatives that were able to prevent the development of bacteria or fungi were used as the minimal inhibitory concentrations (MICs). In order to establish the correctness of the MIC, it was compared with the quantity of control growth that occurred before the incubation process (the original inoculum). The antibiotics gentamycin, ampicillin, chloramphenicol, ciprofloxacin, and norfloxacin served as the standards for determining the antibacterial activity, while nystatin and griseofulvin served as the criteria for determining the antifungal activity [28,40,46].

Conclusions
We designed and developed some methyl 2-((6-chloro-1H-benzo[d]imidazol-2-yl) methylthio)-1,2,3,4-tetrahydro-6-methylpyrimidine-5-carboxylate derivatives, 25-41, as potential DHFR inhibitors. The ADMET profiles of all the created compounds were positive, and some of them even showed reduced binding affinities (in terms of DHFR) compared to the native ligand. In the present investigation, several bacterial and fungal strains (Grampositive bacteria (Staphylococcus aureus and Staphylococcus pyogenes), Gram-negative bacteria (Escherichia coli and Pseudomonas aeruginosa), and fungal (Candida albicans, Aspergillus niger, and Aspergillus clavatus)) were used to evaluate the in vitro antibacterial and antifungal properties of the synthesized compounds. It was observed that 29 and 33 were sensitive against all the bacterial and fungal strains. The results showed that the most promising compounds were those derived from 1,2,3,4-tetrahydropyrimidine-2-thiol with a 6-chloro-2-(chloromethyl)-1H-benzo[d]imidazole moiety, possibly because this moiety plays a vital role in boosting the compounds' antibacterial characteristics. These compounds inhibited the DHFR enzyme with substantial binding affinity. Thus, we reasoned, these compounds may be antimicrobial because they inhibit DHFR. We thus conclude that these compounds have the potential to serve as lead compounds for the creation of further effective antibacterial and antifungal compounds.

Conflicts of Interest:
The authors declare no conflict of interest.