Quantum chemical modelling, molecular docking, synthesis and experimental anti-microbial activity of 1,4-diazepan linked piperidine derivative

Background In this work, we represent synthesis, in silico analysis and biological activity of 1,4 diazepine linked piperidine derivatives (6a–6o). All the derivatives were screened for their anti-microbial activity against gram-positive ( Staphylococcus aureus, Bacillus Subtills, Bacillus megaterium ) and gram-negative ( Escherichia coli, Pseudonymous, Shigella sp.) bacteria. Compounds were synthesized from reaction of tert-butyl 1,4-diazepane-1-carboxylic, butyryl chloride and varied aromatic aldehyde, further characterized by 1 H NMR and LCMS spectral techniques. Result Using ampicillin as a positive control, the synthetic compounds 6a–6o were tested for their in-silico study and experimental anti-microbial activity against gram-positive ( Staphylococcus aureus, Bacillus Subtills, Bacillus megaterium ) and gram-negative ( Escherichia coli, Pseudonymous, Shigella sp.) bacteria. According to in vitro assay compound 6a, compound 6c, compound 6d, compound 6m and compound 6I showed higher activity against all the tested strains. Molecule 6i, compound 6j, compound 6k, compound 6f has good to moderate antibacterial activity. DFT computations were used to optimize the molecular geometry at the B3LYP/6-31G (d, p) theoretical level. The corresponding energy values of molecular orbitals were visualized using optimized geometries. Moreover, Auto Dock Vina 1.2.0 is used to assess molecular docking against two target proteins, Bacillus subtilis (PDB ID: 6UF6) and Protease Vulgaris (PDB ID: 5HXW). The target molecule 6b displayed the best binding energies for both. Additionally, we calculated the ADME for each molecule (6a–6o). Conclusion All fifteen synthesized compounds were screened for their in vitro and in silico analysis. In vitro analysis for anti-microbial activity was carried out against gram-positive ( Staphylococcus aureus, Bacillus Subtills, Bacillus megaterium ) and gram-negative ( Escherichia coli, Pseudonymous, Shigella sp.) bacteria and compound 6a, compound 6c, compound 6d, compound 6m and compound 6I exhibits more potent activity towards all tested strains. Molecular docking is performed against target proteins, L-amino acid deaminase from Proteus Vulgaris and LcpA ligase from Bacillus subtilis , representing the Gram-negative bacterium and Gram-positive bacterium, respectively. Compound 6b showed the highest no. of interaction with protein according to molecular docking. With the advent of innovative techniques like ADME, we select their hit compounds early on and anticipate future pharmacokinetic and pharmacodynamic benefits and drawbacks of these promising therapeutic candidates.

Further, piperidine is a fundamental heterocyclic part of medical chemistry.Piperidine core compounds exhibit pharmaceutical properties like anti-HIV, anticancer.A.R. Zala and colleagues developed piperidinecontaining molecules as antibacterial drugs in 2023 [22].Mahmat Yildiz and colleagues developed piperidine derivatives that act as antimicrobial agents [23].In contrast, Enrico Casalone and colleagues developed 1,4-diazepane derivatives as antimicrobial agents in 2020 [24].Since, both piperidine and 1,4 diazepane show potent antimicrobial activities, in this study we created 1,4 diazepane linked piperidine derivatives and have characterize their antimicrobial efficacy.The derivative molecules have also been investigated using a series of computational chemistry tools, such as the Density Functional Theory, Molecular docking and ADME analysis.Here, we aim to facilitate the current research in designing innovative strategies for

Methods
All the compounds and solvents were purchased from Sigma Aldrich and used without any additional purification.The gradient of MeOH in MDC was created using the elutes and was used for the thin layer chromatography (TLC) method on silica gel plates (60F254, 0.2 mm thick, Merck).In this study, we used a Bruker Advance II 400 NMR spectrometer, used with the internal standard tetra methyl silane (TMS) as a 1 H NMR spectrometer, to perform 1 H NMR spectra in CDCl 3 solution at 400 MHz.Parts per million, or ppm, are used to compute the value in 1 H NMR.
As the mobile phase of the LCMS equipment, which used WATERS to record data, 0.15% formic acid in acetonitrile was used as a mobile phase for the LCMS equipment.

Chemistry
The five-step synthesis process outlined in Scheme 1 was created to produce the desired anti-microbial active ingredient.Commercially available tert-butyl 1,4-diazepane-1-carboxylic (1) intrigues the production route when it is coupled with butyryl chloride in the presence of triethanolamine (TEA) and methylene chloride (DCM) to produce compound 2 [25].Compound 3 is produced by carbamate hydrolysis using compound 2 and 6 N HCl in dioxane to remove the boc protecting group [26].In the presence of TEA, ZnCl 2 , and MeOH, compound 4 was produced by treating compound 3 with commercially available tert-butyl-4-oxopiperidine-1-carboxylate [27].Once more, carbamate hydrolysis is performed to remove the boc protecting group in preparation for the transition from compound 4 to compound 5 [26].With the addition of DIPEA, the carboxylic acid in DMF was reacted with compound 5 in the presence of HATU and generated a high quality desired derivative (6a-o).[28,29].All the used aldehydes and properties of desired derivatives shown in Table 1.

In vitro assay
For the investigation of in vitro anti-microbial activity, the cup plate method [30,31], which is very popular, was used.Gram-positive (Staphylococcus aureus, Bacillus Subtills, Bacillus megaterium and gram-negative (Escherichia coli, Pseudonymous, Shigella sp.) microorganisms, as well as Staphylococcus aureus (Gram +), were used to test in-vitro anti-microbial activity.To carry out this approach, all bacterial cultures were first kept in nutrient broth and then incubated overnight at a temperature of 37 °C.A full hour was given for the incubated molten agar to set and solidify before it was transferred to the sterilized petri plates.Bacterial culture was evenly swabbed on the sterile plates using a cotton swab.On agar media, 6 mm broad bores were constructed using sterilized cork.Each bore was filled with the test compound solution, which had a measured concentration of 1000 µg/ml, using a sterile tip and micropipette.A plate was similarly prepared for the positive control ampicillin, which is taken as standard.The prepared plates were incubated for 24 h at 37 °C.Every bore had a clear zone encircling it after the incubation process, demonstrating that the 1,4-diazepane linked piperidine derivative under test had anti-microbial efficacy.The average diameter of these formed clear zones of inhibition was determined and used to report activity in millimeters.A value for the MIC (minimum inhibitory concentration) was obtained using the liquid dilution method [32], All derivatives MIC values were established for Staphylococcus aureus (Gram+) bacterium.The investigated substances were dissolved in a suitable solvent at a concentration of 50 mg/ml.Similar to this, a conventional ampicillin solution was made at a 50 mg/ml concentration.Bacterial culture inoculum preparation was carried out.In a series of test tubes, 0.2 ml of inoculum is inserted along with a 1 ml solution of the test substance at a specified concentration.In each test tube, 3.8 ml of sterile water was also added.All these test tubes were maintained under observation and incubated for a day in order to detect the existence of turbidity.A similar process was used to screen ampicillin, which is a commonly used medication.MIC values are those where bacterial growth was not seen to occur.

Computational methods
The electronic structures of all 1,4 diazepane linked piperidine derivative molecules (6a-6o) were studied using quantum mechanical calculations.Each molecule was constructed using GaussView 5 software [33] and was subjected to geometry optimization using DFT calculations at the B3LYP/6-31G(d,p) level of theory as implemented in Gaussian 09 quantum chemistry package [34].The calculations were performed in an implicit PCM water solvent model.No symmetrical constraints were imposed during the optimization.Optimized geometries were used to visualize molecular orbitals and their associated energy values using GaussView 5.In particular, the nature and energetics of the lowest unoccupied molecular orbital (LUMO) and highest occupied molecular orbital (HOMO) were characterized.Several other molecular properties were calculated from the HOMO and LUMO energies, namely, the global hardness (η), softness (σ), electronegativity (χ), and electrophilicity index (ω).The following equations are used to compute these properties: Next, the optimized geometry of each derivative molecule was used to perform molecular docking against two target proteins corresponding to the Protease Vulgaris (PDB ID: 5HXW) [35] and Bacillus subtilis (PDB ID: 6UF6) [36].The starting structure for the target proteins were based on their crystal structure deposited in the RCSB protein data bank.Both proteins were prepared for molecular docking by removing any existing ligand and (1) water molecules, followed by charge assignment using MGL Tools 1.5.7.[37] The grid box corresponding to the whole protein was defined for each target.Each derivative molecule was prepared for docking using MGL Tools 1.5.7.AutoDock Vina 1.2.0 [38] was used for docking the derivate molecules against the two target proteins.Ten conformations were generated for each derivative molecule and the most favorable conformation, based on the binding affinity, was selected for further structural analysis.The non-covalent interactions between the protein and docked ligand were analysed and visualized using LigPlot software [39].
Further, the Absorption, Distribution, Metabolism, and Excretion (ADME) analysis was carried out for each derivatized molecule using the SwissADME server [40] The ADME analysis provides determines various physiochemical properties to assess the pharmacodynamic properties of a potential ligand molecule.Various relevant parameters, such as the Lipinski's rule of five violations, number of hydrogen donors, hydrogen acceptors, rotatable bonds, total polar surface area, skin permeability (Log Kp), molar refractivity, gastrointestinal absorption (GI), blood brain barrier (BBB), inhibition of cytochromes P450 isoforms (CYP1A2, CYP2C19, CYP2C9, CYP2D6) were estimated.
Next, we carried out Quantitative Structure-activity relationships (QSAR) to determine the quantitative relationship between the biological activity and the molecule properties of the derivatives.We constructed the QSAR models by performing multiple linear regression analysis, which links selected independent variables with the dependent variable using the following equation:  Where X i are the molecular descriptors, Y is the biological activity, n is number of descriptors, a 0 is the constant and a i are the respective coefficients.

Result
By observing structural alterations brought on by the attachment of various substitutions and their placement on the heterocyclic ring in the generated derivative, the link between structure and activity was investigated (Fig. 2).Because of the varied positions of the aryl group and heterocyclic ring, biological activity was changing.In this study, different substituted phenyl rings linked to piperidine and piperidine directly coupled to 1,4-diazepane were found to have different anti-microbial activities.Distinct substitutions have distinct electrical properties, and these differences are shown in their antimicrobial effectiveness.According to research on SAR, anti-microbial activity can vary depending on the heterocyclic or aryl ring.The heterocyclic ring substitutions of manufactured compounds could either donate electrons or draw them out.When substitutions such halogenated groups at various positions on a heterocyclic ring are made, the anti-microbial action increases.According to Table 2, when phenyl rings are substituted in compounds with different halogenated groups, the chloro group exhibits more activity than the bromo group.The enhanced antimicrobial activity was found to be more in both 6c and 6h, due to the heterocyclic moiety in the synthetic derivatives with alkoxy phenyl ring in 6h and the heterocyclic moiety on the heterocyclic ring with -Br and -F in 6c.Higher antibacterial inhibitory action was demonstrated in 6m by the phenyl ring with the chloro group in the ortho position.As opposed to the meta-substituted 6e and the amine-aryl ring with a bromine substitution in 6g, the phenyl ring with a bromine and methyl group in 6f shown stronger inhibitory activity.Higher antibacterial inhibitory action was demonstrated by compound 6a's phenyl ring with the nitro group in the ortho position as compared to compound 6d's phenyl ring with the electron-donating and withdrawing groups replaced.In contrast, 6l had reduced activity due to the presence of an electron releasing group at the o-position.The methoxy group, a weak electron donor, was shown to be less active than other groups in position 4 of the phenyl ring in compound 6k.Other substituted derivatives 6i, 6b, and 6j displayed modest activity, while 6o was found to be more effective than regular ampicillin against Staphylococcus aureus and E. coli.
Table 2 shows that the majority of the substances tested had varying inhibitory effects on the growth of the bacterial strains that were put to the test, according to the results of the study.Because the compound 6a, the compound 6d, and the compound 6m, compound 6l both possess electron withdrawing and electron releasing groups, these compounds showed a high level of inhibitory activity against the tested microbial strains.Comparatively to the reference drug Ampicillin, compound 6c substituted with -Br and -F on its heterocyclic ring as well as compound 6h containing heterocyclic rings attached to methoxybenzene were found to exhibit superior activity against all Gram-positive and Gramnegative microbial strains.Even though compound 6o showed no effect against B. subtilis, Bacilius magneterium, Pseudomonas spp., and Shigella spp., compound 6j lost its antibacterial efficacy due to the absence of an aryl ring substitution.Due to the presence of an electron

DFT calculations
All 1,4 diazepane linked piperidine derivative molecules were optimized using the DFT methods at the B3LYP/6-31G(d,p) level of theory.The optimized geometries are shown in Fig. 4 and the various electronic properties are presented in Table 3. First, we compared the energy gap (E HOMO -E LUMO ), which is an important parameter to assess the thermal stability and the chemical reactivity of a molecule.The E HOMO and E LUMO energy gap were computed to be negative for all derivative molecules, indicating their stability and in the following order: 6b > 6k > 6l > 6j > 6f > 6e > 6m > 6a > 6c > 6g > 6n > 6o > 6i > 6h > 6d.
Further, Frontier Molecular Orbital analysis illustrate that for all molecules, the lowest occupied molecular orbital (LUMO) orbital is localized on the π orbitals of the phenyl ring (Fig. 5).Whereas the highest occupied molecular orbital (HOMO) orbital is localized on the diazepane and piperidine binding site.We also computed additional electronic and structural parameters, such as electronegativity (χ), global hardness (η), global softness (σ) and global electrophilicity index (ω), to ascertain the biological activity of the derivative molecules (Table 3).The large χ and ω for all compounds indicate their excellent bioactivity.

Molecular docking
The derivative molecules were next subjected to molecular docking to assess their antibacterial potential.To this end, we used the geometry optimized derivative molecules to perform molecular docking against two target proteins, L-amino acid deaminase from Proteus Vulgaris and LcpA ligase from Bacillus subtilis, representing the Gram-negative bacterium and Gram-positive bacterium, respectively.The relative stability of the target proteinligand complex was determined by their binding affinities Table 4 and their intermolecular interactions with the target proteins (Additional file 1: Figures S21-S22).Molecules 6b showed the largest binding affinity for both target proteins (Fig. 6).Further analysis of the proteinligand complex revealed a higher number of interactions in the case of 6b (Fig. 7).The derivate 6b formed H-bonds involving residues Tyr286 and Tyr284 of the L-amino acid deaminase from Proteus Vulgaris.Similarly, it formed H-bonds with Asn353 and Gln350 residues of LcpA ligase from Bacillus subtilis.These H-bonds, together with several hydrophobic interactions, provide higher binding strength to the complex of 6b with target proteins.

ADME calculations
Next, we performed the ADME calculations for the derivative molecules using the SwissADME server to characterize their drug-likeness and bioavailability.The ADME analysis showed that none of the molecules violated any of the five Lipinski rule.(Table 5) Therefore, these molecules have good potential to be developed as orally active drugs.The TPSA value, closely related to the bioavailability, for the molecules were observed in the range of 56 Å 2 to 102 Å 2 , which is well below the limit of 140 Å 2 .Similarly, the number of rotatable bonds (NRB) for all molecules is less than the limit of 10, except for 6i, suggesting that the molecules are conformationally stable.Moreover, the low skin permeability value (Log K p ) for all molecules indicates low level of skin permeation (Table 6).All derivative molecules showed high level of gastrointestinal (GI) absorption.The molecule's hazardous or adverse effects also depend on its suppression of cytochromes P450 isoforms (CYP1A2, CYP2C19, CYP2C9, and CYP2D6).SwissADME predictions showed that all the derivative molecules show inhibition propensity against one or more of these isoforms.
Considering all the ADME predictions as well as the binding affinities, molecule 6b appear as the most effective lead as an antibacterial agent.While it is noted that other derivative molecules, such as 6n, showed high binding affinity and similar binding mode as 6b, the latter is more favourable owing to its higher tendency to form H-bonds with the nearby receptor residues.
Among the 15 molecular descriptors, seven were selected for the regression analysis based on their correlation coefficient values.The seven descriptors are the following: Log P, Log S, NHA, η, molecular weight (MW), TPSA, and Log Kp.The resulting QSAR model based on the regression analysis is given by the following equation: Where N = 15, R = 0.87, determination coefficient (R 2 ) = 0.77, mean square error (MSE) = 1.77, statistical confidence degree (F) = 5.68.
The high values of R2 and F indicate that the derived QSAR model is acceptable, and that the biological activity is linear correlated with these descriptors (Additional file 1: Figure S23).These descriptors can be used to predict inhibitory activity of new compounds based on the structure of 1,4 diazepane linked piperidine derivatives.

General procedure of compound (6a-o)
The carboxylic acid (1 equivalent) was mixed in DMF (10 volumes) in a reaction vessel.Next, the resulting mixture was cooled to absolute zero.After adding HATU (1.5 equivalents), the reaction mixture was stirred at 0 °C for 10 min.Ethyl 4-(piperidin-4-yl)-1,4-diazepane-1-carboxylate hydrochloride salt (5) (1.1 equivalents) and DIPEA (3 equivalents) were then added to the reaction mixture.The end product was stirred at room temperature for six hours.The reaction mixture was then dumped into ice-cold water after it had finished.The resultant residue was divided between 500 mL of H 2 O and 500 mL of EtOAc.EtOAc (250 mL) was used to extract the aqueous layer further.Na 2 SO 4 was used to mix and dry the organic layers.The crude product was then produced after the solvent was removed under vacuum.To produce the final chemicals 6a-o, column chromatography was used to purify the crude product.(Supplementary File).

Figure 3
depicts the graphical representation of antibacterial activity.

Fig. 5
Fig.5 The HOMO and LUMO orbitals for selected derivative molecules (6b and 6c) as calculated at the B3LYP/6-31G(d,p) level of theory

Fig. 6 Fig. 7
Fig.6 Comparison of the calculated docking affinities (in kcal/mol) of all derivative molecules and ampicillin against the target proteins from Bacillus subtilis (PDB ID: 6UF6), and Proteus Vulgaris (PDB ID: 5HXW)

Table 1
Properties of compounds

Table 2
Antibacterial activity of tested compounds as a zone of inhibition in MIC(μg/mL) of synthesized compound

Table 4
Calculated docking affinities (in kcal/mol) of the derivative molecules and ampicillin against the target proteins from Bacillus subtilis (PDB ID: 6UF6), and Proteus Vulgaris (PDB ID: 5HXW)

Table 5
Drug-likeness predictions for the derivative molecules as computed using SwissADMEThe number of rotatable bonds (NRB), number of hydrogen donors (NHD), number of hydrogen acceptors (NHA), total polar surface area (TPSA) in Å 2 , LogP value and Lipinski's rule of five violation (LV), and Molar Refractivity (MR) are reported

Table 6
Absorption, Distribution, Metabolism, and Excretion (ADME) analysis for the derivative molecules as computed using SwissADME