3.3 Evaluation of ADMET properties
ADMET analysis was conducted on compounds 4, 7, 25, and 26 to evaluate their ADME and toxicological properties and to compare the same with those of the reference drugs (Miltefosine and Pentamidine). The predicted ADME and toxicological properties were presented in Table 5.
Table 5
Predicted ADME and toxicological properties of the selected diselenides and reference drugs
ADMET properties
|
Reference drugs
|
Selected diselenides
|
Milt
|
Pent
|
4
|
7
|
25
|
26
|
Absorption
|
Human Intestinal Absorption (%)
|
92.02
|
77.04
|
76.88
|
77.63
|
64.94
|
93.07
|
Skin permeability
|
-2.73
|
-2.87
|
-2.74
|
-2.74
|
-2.74
|
-2.74
|
P-glycoprotein substrate
|
-
|
-
|
-
|
-
|
-
|
-
|
P-glycoprotein inhibitor
|
-
|
-
|
+
|
+
|
+
|
+
|
|
Distribution
|
BBB permeability (LogBB)
|
-0.173
|
-0.905
|
-1.074
|
-1.122
|
-0.712
|
-0.148
|
CNS permeability (LogPS)
|
-3.191
|
-2.941
|
-2.113
|
-2.671
|
-2.705
|
-2.957
|
|
Metabolism
|
CYP2D6 substrate
|
-
|
-
|
+
|
+
|
+
|
+
|
CYP3A4 substrate
|
+
|
-
|
+
|
+
|
+
|
+
|
CYP2D6 inhibitor
|
-
|
-
|
-
|
-
|
-
|
-
|
CYP3A4 inhibitor
|
-
|
-
|
+
|
+
|
+
|
+
|
|
Excretion
|
Total clearance (Log ml/min/kg)
|
0.112
|
0.850
|
2.096
|
2.515
|
2.921
|
2.908
|
Renal OCT2 substrate
|
-
|
+
|
-
|
-
|
-
|
-
|
|
Toxicity
|
AMES mutagenesis
|
-
|
-
|
-
|
-
|
-
|
-
|
MRTD (Log mg/kg/day)
|
0.099
|
0.522
|
0.480
|
0.496
|
0.584
|
0.577
|
hERG inhibitor
|
-
|
-
|
-
|
-
|
-
|
-
|
Hepatotoxicity
|
-
|
+
|
+
|
-
|
+
|
-
|
Skin sensitization
|
-
|
-
|
-
|
-
|
-
|
-
|
Eye irritation
|
+
|
-
|
-
|
-
|
-
|
-
|
Carcinogenicity
|
-
|
-
|
-
|
-
|
-
|
-
|
Nephrotoxicity
|
+
|
-
|
+
|
-
|
+
|
+
|
The estimated ADME and toxicological properties reported in Table 5 showed good HIA for all the studied compounds (greater than 30% threshold). In transdermal drug delivery development, evaluation of skin permeability is crucial (Ugbe et al., 2023a). The various analogues showed a skin permeability constant (Log Kp) of less than − 2.50, indicating good skin permeation. The logarithmic ratio of brain to plasma drug concentration (LogBB) and the blood-brain permeability surface area product (LogPS) is important to evaluate the BBB and CNS permeability respectively (Thai et al., 2020). As a result, all the tested compounds showed LogBB of less than 0.3 and LogPS of less than − 2, indicating they do not readily permeate the BBB and CNS respectively. Also, the p-glycoprotein enzyme which helps to eliminate toxins and xenobiotics from the cell was predicted, in which all the compounds were substrates of the enzyme but only the selected diselenides inhibit the enzyme, connoting that these analogues when taken into the human system may easily not be effluated by the enzyme p-glycoprotein. Two important isoforms of the cytochrome P450 enzymes; CYP-3A4 and CYP-2D6 were predicted. These enzymes are important in the body to facilitate the metabolism and excretion of drugs (Zhou & Lauschke, 2022). All of the tested diselenides are substrates of both enzymes. Pentamidine is neither a substrate nor an inhibitor of both enzymes, while Miltefosine is a substrate of CYP-3A4 only. However, all the diselenides are inhibitors of CYP-3A4. This result connotes a good drug metabolism in the body for the selected diselenides than the two reference compounds. Total clearance provides the extent of drug elimination from the body (Talevi & Bellera, 2021). The values of the predicted total clearance for all tested compounds showed good clearance. For the disposition and renal clearance of drug molecules and endogenous compounds, the renal OCT is crucial. Notwithstanding, OCT 2 is reported to be associated with unwanted drug-drug interaction, making it important to predict whether a drug molecule is a substrate or inhibitor of OCT 2 (Galeano et al., 2020). Consequently, all tested compounds except the reference drug (Pentamidine) are non-OCT 2 substrates. Furthermore, some toxicity indices were predicted to provide insight into the safety profiles of newly designed analogues. All the compounds showed a negative AMES toxicity, connoting no associated risk of mutagenicity. MRTD was predicted for the various compounds; MRTD ≤ 0.477 log (mg/kg/day) is considered low, while a value greater than 0.477 log (mg/kg/day) is high. Therefore, all tested compounds except Miltefosine have high MRTD value. Inhibitors of hERG could be responsible for the acquired long QT syndrome, causing irregular heartbeats (Tschirhart et al., 2019). None of the tested molecules was implicated. Additionally, only 4, 25, and Pentamidine are hepatotoxic. Also, all the tested compounds showed negative skin sensitization, eye irritation, and carcinogenicity except Miltefosine with positive eye irritation. Lastly, only 7 and Pentamidine are non-nephrotoxic.
Based on the drug-likeness, ADME, and toxicological analyses conducted, the selected diselenides (Lead compounds) showed good pharmacokinetic properties and compared favourably with both reference drugs.
3.4 Pharmacological interaction study
The 2-D and 3-D views of the binding interactions of the lead compounds (4, 7, 25, and 26) with OASS (3SPX) as obtained from MOE are shown in Figs. 1 and 2 respectively, while Figs. 3 and 4 respectively showed the 2-D and 3-D views of their molecular interactions with PdxK (6K91). Also, the various protein-ligand interactions in terms of amino acid residues and interaction distances were summarized in Tables 6 and 7 for OASS and PdxK respectively.
The 2-D and 3-D binding interaction images were obtained to provide insight into the binding mode of selected diselenides in interaction with OASS and PdxK (Figs. 1–4). Our results showed that the studied compounds interacted more strongly with OASS than PdxK. Compound 25 is the most interactive ligand with OASS through four hydrogen-bond (H-bond) donors with MET 130, PRO 233, THR 83, and THR 83, four H-bond acceptors with PRO 225, SER 274, SER 302, and GLY 230, and a π-H interaction with ASN 82 through its 6-ring (Table 6). Furthermore, compound 4, a ligand with the highest affinity for PdxK, established five H-bond donors with GLU 154, ASP 125, and THR 229, and two H-bond acceptors with GLN 258 and SER 188 (Table 7).
Table 6
Predicted binding interaction profiles of selected diselenides with OASS (3SPX)
Comp. ID
|
MolDock
|
Ligand
|
Amino Acid
|
Interaction
|
Distance (Å)
|
4
|
-190.03
|
SE1 1
|
O SER 78 (A)
|
H-donor
|
3.85
|
N2 13
|
O HIS 226 (A)
|
H-donor
|
2.92
|
N4 39
|
OG1 THR 83 (A)
|
H-donor
|
2.82
|
N5 41
|
O GLY 230 (A)
|
H-donor
|
2.88
|
N3 23
|
N THR 311 (A)
|
H-acceptor
|
3.09
|
N3 23
|
OG1 THR 311 (A)
|
H-acceptor
|
2.93
|
N6 51
|
N LEU 191 (A)
|
H-acceptor
|
2.99
|
6-ring
|
CA GLY 230 (A)
|
pi-H
|
3.76
|
7
|
-200.59
|
SE1 1
|
OE1 GLN 229 (A)
|
H-donor
|
3.56
|
SE2 2
|
SD MET 106 (A)
|
H-donor
|
3.64
|
N3 45
|
O GLY 232 (A)
|
H-donor
|
3.33
|
N4 47
|
O GLY 232 (A)
|
H-donor
|
2.99
|
O2 33
|
ND2 ASN 82 (A)
|
H-acceptor
|
2.88
|
6-ring
|
N SER 80 (A)
|
pi-H
|
3.88
|
TYR-176
|
OE1 GLN 229 (A)
|
H-donor
|
3.56
|
25
|
-203.02
|
SE1 1
|
O MET 130 (A)
|
H-donor
|
3.28
|
SE2 2
|
O PRO 233 (A)
|
H-donor
|
3.63
|
N4 39
|
OG1 THR 83 (A)
|
H-donor
|
3.14
|
N5 41
|
OG1 THR 83 (A)
|
H-donor
|
2.97
|
SE3 27
|
CA PRO 225 (A)
|
H-acceptor
|
4.16
|
N6 51
|
OG SER 274 (A)
|
H-acceptor
|
2.79
|
N6 51
|
OG SER 302 (A)
|
H-acceptor
|
2.92
|
SE4 55
|
CA GLY 230 (A)
|
H-acceptor
|
3.80
|
6-ring
|
CB ASN 82 (A)
|
pi-H
|
3.94
|
26
|
-199.51
|
SE1 1
|
SD MET 130 (A)
|
H-donor
|
4.03
|
N2 13
|
O HIS 226 (A)
|
H-donor
|
3.01
|
SE3 30
|
SD MET 106 (A)
|
H-donor
|
3.85
|
N4 44
|
OG1 THR 83 (A)
|
H-donor
|
2.90
|
SE4 61
|
ND2 ASN 82 (A)
|
H-acceptor
|
4.01
|
6-ring
|
N GLY 230 (A)
|
pi-H
|
3.91
|
In general, the various compounds fit well into the target site cavities of both target receptors making very close contacts with good number of the amino acid residues giving rise to hydrogen bonding interactions, the most important interaction type in any drug-receptor binding. Imberty et al. (1991), classified hydrogen bonding interaction distances as either strong or weak based on the distance between hydrogen donor and hydrogen acceptor (dis(D−A)) as follows; 2.5Å ˂ dis(D−A) ˂ 3.1Å (strong hydrogen bond) and 3.1Å ˂ dis(D−A) ˂ 3.55Å (weak hydrogen bond). Our results showed a mixed strong and weak H-bonding interactions (Tables 6 and 7). Therefore, the studied compounds have demonstrated the potential to arrest both target receptors (OASS and PdxK), absolutely necessary factors governing many activities which are essential for viability of the leishmania organisms.
Table 7
Predicted binding interaction profiles of selected diselenides with PdxK (6K91)
Comp. ID
|
MolDock
|
Ligand
|
Amino Acid
|
Interaction
|
Distance (Å)
|
4
|
-198.18
|
SE1 1
|
OE2 GLU 154
|
H-donor
|
3.87
|
N1 11
|
OD2 ASP 125
|
H-donor
|
2.75
|
N2 13
|
OD2 ASP 125
|
H-donor
|
2.81
|
N4 39
|
O THR 229
|
H-donor
|
2.83
|
N5 41
|
O THR 229
|
H-donor
|
2.77
|
N6 51
|
NE2 GLN 258
|
H-acceptor
|
2.87
|
O2 55
|
OG SER 188
|
H-acceptor
|
2.65
|
7
|
-193.96
|
SE1 1
|
O ALA 253
|
H-donor
|
3.82
|
SE2 2
|
O MET 254
|
H-donor
|
3.58
|
N4 47
|
OD2 ASP 124
|
H-donor
|
2.81
|
O2 33
|
OG1 THR 229
|
H-acceptor
|
2.70
|
25
|
-191.51
|
SE2 2
|
OD2 ASP 231
|
H-donor
|
3.21
|
N1 11
|
OD2 ASP 124
|
H-donor
|
3.38
|
N4 39
|
OD1 ASP 124
|
H-donor
|
2.85
|
N6 51
|
NE2 GLN 258
|
H-acceptor
|
2.97
|
SE4 55
|
N TYR 226
|
H-acceptor
|
4.00
|
26
|
-185.5
|
SE1 1
|
OD2 ASP 231
|
H-donor
|
3.29
|
N1 11
|
OD2 ASP 124
|
H-donor
|
2.72
|
N2 13
|
OG1 THR 229
|
H-donor
|
3.07
|
N3 42
|
OG SER 47
|
H-donor
|
2.91
|
6-ring
|
OH TYR 129
|
pi-H
|
3.90
|
3.6 Molecular dynamic simulation and MM/GBSA analysis
The results of the docking investigation showed interactions between the studied ligands and the target proteins (OASS and PdxK). Notwithstanding, these results were not enough to confirm whether the protein-ligand interactions could be stable or not as docking only pictures protein as a rigid molecule while ignoring the dynamic nature of protein-ligand interactions (Alameen et al., 2022). Consequently, to complement the docking analysis, therefore, MD simulation was performed on the protein-ligand complexes of the selected diselenides to ascertain their interactions stability in a dynamically perturbed system.
3.6.1 Stability of protein-ligand complexes
The root mean square deviation (RMSD) and root mean square fluctuation (RMSF) are the two most important qualitative indicators of the stability of protein-ligand complexes (Eltayb et al., 2023). RMSD values will indicate how structures and parts of structures change over time as compared to the starting point. A large RMSD value shows a great deviation in structural changes compared to the structure at the starting point and thus indicates less stability of the complex. RMSD is typically plotted vs. time (Abdalla et al., 2022). RMSF on the other hand is a calculation of individual residue flexibility, or how much a particular residue moves (fluctuates) during a simulation. RMSF per residue is typically plotted vs. residue number and can indicate structurally which amino acids in a protein contribute the most to a molecular motion (Alameen et al., 2022). The various RMSD plots of the studied complexes were presented in Figs. 6 and 7 for 3SPX and 6K91 respectively, while Figs. 8 and 9 showed their RMSF plots respectively. During a trajectory, the presence of widespread fluctuation signifies conformational changes in protein-ligand interactions. While RMSD values show the best stability of the compounds, RMSF values indicate the fluctuations that result in the compactness of the complexes (Ononamadu et al., 2021).
The results of the present study (Figs. 6–9) showed a very stable system for the diselenide molecules especially 7_3SPX, 26_3SPX, 7_6K91, 25_6K91, and 26_6K91, indicating that compounds 7 and 26 interact more stably with both receptors. Fewer fluctuations were more associated with 7 and 26 than 4 and 25 in terms of their interactions with OASS (3SPX), while 7, 25, 26 recorded relatively fewer fluctuations in their interactions with PdxK (6K91). For 7_3SPX (Fig. 6C), a major fluctuation in RMSD was observed from 3.2 Å to 5.8 Å within the first 30 ns and tends to reduce the fluctuations to reach equilibration at an average RMSD of 5.2 Å from about 40 ns. The RMSD profile of 26_3SPX (Fig. 6E) showed a major fluctuation between 3.2 Å and 6.4 Å only during the first 10 ns of trajectory before reaching equilibration where fluctuation was stabilized with an average RMSD of 4.8 Å throughout the simulation. The profiles of 4 and 25 witnessed large fluctuations of up to 10.5 Å, but one interesting fact about 25_3SPX (Fig. 6D) is that it reached equilibration at 50 ns and became very stable maintaining the lowest RMSD value of 3.0 Å.
For 7_6K91 (Fig. 7C), a major fluctuation in RMSD was observed from 6.4 Å to 7.2 Å within the first 30 ns and tends to reduce the fluctuations to reach equilibration at an average RMSD of 5.6 Å from about 35 ns. Similarly, 25_6K91 (Fig. 7D) maintained a very stable fluctuation in RMSD between 7 Å and 8 Å from less than 5 n throughout the trajectory. The RMSD profile of 26_3SPX (Fig. 7E) showed a rapid fluctuation between 1.6 Å and 4.0 Å only during the first 10 ns of trajectory before reaching equilibration where fluctuation was stabilized with an average RMSD of 4.8 Å throughout the simulation. On the contrary, the protein RMSD showed stable fluctuation between 1.0 Å and 2.2 Å for 6K91 and between 1.6 Å and 4.8 Å for 3SPX. Therefore, the RMSD profiles of the selected diselenides showed their stability in complex with both receptors with compound 4 showing the least stability, while also comparing very favourably with those of the control (co-crystal) (Fig. 6A and Fig. 7A).
The RMSF results conformed to those of the RMSD trajectories. The RMSF values were kept within the range of 0.5 Å – 4.5 Å for 26, and 0.6 Å − 5.4 Å for 7, 0.6 Å – 4.8 Å for 25, and 0.8 Å – 5.6 Å or 4 (Fig. 8). As shown in Fig. 9, The RMSF values were kept within the range of 0.5 Å – 4.5 Å for 7, 0.4 Å – 3.6 Å for 26, 0.4–3.2 Å for 25, and 0.4–3.2 Å for 4. The average fluctuation in protein residues is greatest in the complexes of compound 4, indicating the least stability in complex with the target receptors. In general, the overall stability of the various complexes based on the RMSD and RMSF analyses was good, with 26 and 7 been outstanding. Also, the profile of compound 26 indicates more stability than that of the co-crystal in both in their interaction with both receptors.
3.6.2 Binding free energy (MM/GBSA) analysis
The binding free energy (∆G) of interaction also provides insight into the favourability and stability of the protein-ligand complexes (Abdalla et al., 2022). In the current study, the MM/GBSA method was used to compute the free energies of interactions, and the results were tabulated in Table 9. MM/GBSA score is an aggregate score of the macromolecular generalized born and surface area solvation score (Sahakyan, 2021). It measures the amount of free energy involved in a particular protein-ligand interaction set (Ugbe et al., 2023c). The higher the free energy (i.e. more negative energy value), the better the interaction profile and stability of the complex. From Table 9, it was observed that all the complexes have negative binding free energies of interactions. This further corroborated the favourability of the interactions between the compounds and the receptor. Also, notwithstanding the relatively greater RMSD fluctuations associated with compound 25, it interacted with 3SPX with the highest value of ∆Gbind (-88.35 kcal/mol). This may be attributed to the fact that 25_3SPX equilibrated and maintained stable fluctuations at the lowest average RMSD value of 3.0 Å. However, there is a clear correlation between the RMSD values and ∆Gbind for compound 4, as it has the least ∆Gbind value of -69.73 kcal/mol, which further corroborates the fact that compound 4 formed the least stable complex with 3SPX. The values of ∆Gbind for 3SPX complexes follow the order; 25 (-88.35 kcal/mol) > 7 (-78.20 kcal/mol) > 26 (-77.09 kcal/mol) > 26 (-69.73 kcal/mol), which is in consonant with the order of binding affinities provided by the docking study. For complexes of 6K91 on the other hand, the values of ∆Gbind follow the order; 4 (-90.29 kcal/mol) > 26 (-87.14 kcal/mol) > 25 (-86.14 kcal/mol) > 7 (-74.37 kcal/mol). Compound 4 which showed a larger fluctuations in RMSD has the highest value of ∆Gbind of -90.29 kcal/mol. This may be attributed to the fact that 4_6K91 could not reach equilibration in 100 ns of simulation time, and hence the observed large fluctuations.
Table 9
Binding energies (MM/GBSA) of the complexes of OASS and PdxK with selected diselenides
Compounds
|
MM-GBSA ∆G Bind (kcal/mol)
|
MM-GBSA ∆G Bind Coulomb (kcal/mol)
|
MM-GBSA ∆G Bind Covalent (kcal/mol)
|
MM-GBSA ∆G Bind Hbond (kcal/mol)
|
MM-GBSA ∆G Bind Lipo (kcal/mol)
|
MM-GBSA ∆G Bind Packing (kcal/mol)
|
MM-GBSA ∆G Bind Solv GB (kcal/mol)
|
MM-GBSA ∆G Bind vdW (kcal/mol)
|
|
3SPX
|
4
|
-69.73
|
-18.75
|
2.34
|
-0.6
|
-21.27
|
-3.46
|
24.88
|
-52.89
|
7
|
-78.2
|
-26.48
|
-2.53
|
-1.64
|
-20.61
|
-1.533
|
27.92
|
-53.33
|
25
|
-88.35
|
-14.79
|
-0.76
|
-1.8
|
-21.18
|
-3.33
|
17.35
|
-63.84
|
26
|
-77.09
|
-9.42
|
-2.41
|
-0.91
|
-18.97
|
-5.02
|
24.66
|
-65.01
|
|
3SPX
|
4
|
-90.29
|
-24.67
|
2.01
|
-2.04
|
-29.24
|
-2.28
|
37.04
|
-71.11
|
7
|
-74.37
|
-19.18
|
-2.65
|
-1.05
|
-23.85
|
-3.98
|
36.11
|
-59.78
|
25
|
-86.54
|
-22.06
|
-0.55
|
-0.24
|
-25.22
|
-5.05
|
41.01
|
-74.44
|
26
|
-87.14
|
-16.7
|
2.5
|
-1.3
|
-31.47
|
-3.33
|
37.1
|
-73.95
|
3.6.3 Protein and ligand properties from MD simulation analysis
Further analyses involving ligand properties were carried out such as RMSD, radius of gyration (rGyr), molecular surface area (MolSA), solvent accessible surface area (SASA), and Polar Surface Area (PSA), with obtained results presented in Figs. 10 and 11 for 3SPX and 6K91 respectively.
In Figs. 10 and 11, the Y-axis represents the values (in Å) for RMSD and rGyr and Å2 for MolSA, SASA, and PSA, while the X-axis represents the simulation time in ns. These ligand properties provide information about the ligand’s behavior in the binding pocket of the receptor (Mitra et al., 2023). From Fig. 10, the RMSD values of the various compounds ranged between 0.0 and 2.4 Å with equilibrium at 1.6 Å for compound 4, 0.0 and 3.0 Å with equilibrium at 2.5 Å for compound 7, 0.0 and 4.5 Å with equilibrium at 1.5 Å for compound 25, and 0.0 and 3.0 Å with equilibrium at 2.8 Å for compound 26. Similarly, for interactions involving 6K91 the RMSD values ranged between 0.0 and 3.0 Å for compound 4, 0.0 and 4.5 Å with equilibrium at 3.0 Å for compound 7, 0.0 and 3.0 Å with equilibrium at 2.5 Å for compound 25, and 0.0 and 3.0 Å with equilibrium at 2.0 Å for compound 26. Therefore, the overall RMSD values of the complexes suggest great stability and good binding poses.
The rGyr provides information on the protein compactness (Abdalla & Rabie, 2023). The lower the rGyr, the higher the compactness and the better the stability. The ligands showed only slight fluctuations within a small limit and then reached equilibrium. The rGyr values were 6.0–7.5 Å (equilibrium at 7.0 Å), 4.2–6.0 Å (equilibrium at 4.8 Å), 5.6–8.0 Å (equilibrium at 6.4 Å), and 6.6–7.8 Å (equilibrium at 6.6 Å) respectively for 4, 7, 25, and 26 in their complexes with OASS (Fig. 10B-E). Similarly, for interactions with PdxK, the rGyr values were 6.6–8.4 Å (equilibrium not attained), 5.0–7.0 Å (equilibrium at 6.0 Å), 6.5–7.5 Å (equilibrium at 7.5 Å), and 6.4–7.6 Å (equilibrium at 7.2 Å) respectively for 4, 7, 25, and 26 (Fig. 11B-E). These values were steady during the simulation, an indication that the protein folding was stable and compact.
MolSA is also linked to the stability of the protein–ligand complexes. A high MolSA value indicates an unstable protein–ligand complex, whereas a lower MolSA value indicates a comparatively stable complex (Ferdausi et al., 2022). For interaction with OASS (Fig. 10B-E), the fluctuation in MolSA is least for compound 4 (480–495 Å2), then compound 26 (520–535 Å2), and greatest in 7 (440–520 Å2). Also, for interaction with PdxK (Fig. 11B-E), the fluctuation in MolSA is least for compound 26 (516–528 Å2), then compound 4 (474–492 Å2), and greatest in 7 (460–520 Å2). In general, the range of fluctuations in MolSA suggests the stability of the various complexes. A complex with high SASA values indicates high solvent accessibility, leading to increased instability, while a complex with a low SASA value indicates a relative stability (Ferdausi et al., 2022). For ligand interactions with OASS, the values of the SASA fluctuate between 80 and 320 Å2 for 4, 150 and 450 Å2 for 7, 120 and 300 Å2 for 25, and between 160 and 300 Å2 for 26 (Fig. 10B-E). Similarly for ligand interactions with PdxK, the values of the SASA fluctuate between 120 and 300 Å2 for 4, 100 and 300 Å2 for 7, 60 and 180 Å2 for 25, and between 60 and 340 Å2 for 26 (Fig. 11B-E). The PSA is the SASA of a ligand contributed only by oxygen and nitrogen atoms. If the PSA of a drug is more than 40 or less than 90Å, it may cross the blood–brain barrier (Ferdausi et al., 2022). In this study, the PSA has its values as follows; 225–235 Å2 for 4, 120–160 Å2 for 7, 152–168 Å2 for 25, and 72 to 90 Å2 for 26 in their interactions with OASS (Fig. 10B-E). Also, for interactions with PdxK (Fig. 11B-E), the PSA values range from 225–235 Å2 for 4, 120–150 Å2 for 7, 150–162 Å2 for 25, and 64 to 80 Å2 for 26. This showed that only compound 26 can trans-pass the BBB. In general, there is no widespread variation of ligand properties among the ligands studied with the calculated values suggesting their stability within the binding pocket of both receptors.
3.6.4 Protein-ligand contacts and bonds interactions
When drug molecules and/or chemical compounds interact with proteins, bonds (covalent or no-covalent) are formed. The most important bonds in drug-receptor interaction are hydrogen bonds (H-bond), hydrophobic, ionic, Vander Waals, and water bridges (solvent bond). H-bond and ionic bonds constitute polar interactions, while hydrophobic interactions are non-polar resulting from non-polar parts of the protein and the ligands (Klebe, 2013). Based on the binding affinities of interactions obtained from docking, the various compounds were further inspected for interactions with OASS and PdxK in terms of H-bonding, hydrophobic, and other interaction types as shown in Figs. 12 and 13 respectively.
For interactions involving OASS (Fig. 12), H-bond interactions were visible between compound 4 and the following residues; ASN-82, TYR-83, GLY-154, GLN-229, GLY-230, GLY-234, and ASN-278. For 7, the H-bond residues include SER-78, ASN-82, MET-130, THR-187, GLN-229, GLY-230, GLY-232, and GLY-234. For 25, the H-bond residues are SER-80, THR-83, HIS-226, PRO-233, GLY-234, and SER-294, while ASN-82, GLN-152, HIS-226, and GLY-229 were the residues in H-bond interactions with 26. Also for interactions involving PdxK (Fig. 13), H-bond interactions were visible between compound 4 and the following residues; HIS-46, SER-47, ASP-124, ASP-125, TYR-129, ASN-181, ARG-225, THR-229, GLN-258, and ARG-278. For compound 7, the H-bond residues include ASN-181, TYR-182, LYS-187, SER-188, ALA-190, PRO-220, TYR-221, HIS-222, GLU-223, GLY-224, and TYR-226. For 25, the H-bond residues are SER-12, SER-47, ASN-87, SER-188, HIS-222, THR-229, and GLN-258, while SER-47, VAL-121, GLY-123, CYS-130, ASN-151, SER-188, HIS-222, THR-229, GLN-258, and ARG-278 were the residues in H-bond interactions with 26.
For hydrophobic interactions involving OASS (Fig. 12), the ligands were bound to the following residues: VAL-50, MET-106, MET-130, PRO-225, ILE-231, PRO-233, PHE-235, and PRO-300 for compound 4; MET-130, VAL-134, PHE-153, PRO-233, and PHE-235 for compound 7; MET-130, LYS-131, PHE-153, PRO-223, PRO-225, HIS-226, and PHE-235 for 25; and VAL-50, LEU-54, PHE-153, HIS-226, PRO-233, PHE-273, and TYR-307 for compound 26. Also for hydrophobic interactions involving PdxK (Fig. 13), the ligands were bound to the following residues: HIS-46, TYR-49, VAL-121, TYR-129, LEU-196, HIS-222, PHE-233, LEU-257, and ARG-278 for compound 4; TYR-152, PHE-153, ALA-190, LEU-198, VAL-218, TYR-221, HIS-222, TYR-226, LEU-257, ILE-261, and ILE-265 for compound 7; VAL-19, LEU-43, ILE-52, TYR-85, VAL-121, TYR-129, TYR-152, PHE-153, LEU-198, HIS-222, LEU-257, and ILE-261 for compound 25; and VAL-19, ILE-52, TYR-85, VAL-121, TYR-129, LEU-198, VAL-219, TYR-221, LEU-257, and ILE-262 for compound 26.
Only limited ionic interactions were observed only in compound 4 in complex with OASS, involving HIS-226, GLN-229, and GLY-230. Additionally, a good number of solvents interactions (Water Bridge) were observed in all the complexes. Water Bridge is important as the surface residues of the protein form the bond thereby enhancing the transfer of electrons between the ligands and the receptors (Alameen et al., 2022). The overall results indicate that all the studied compounds showed interesting interactions with the various amino acid residues of OASS and PdxK in diverse interaction fractions.
To further evaluate the stability of these bonds, the protein-ligand contacts within the simulation trajectory were plotted and shown in Figs. 14 and 15 for interactions with OASS and PdxK respectively. The number of contacts fluctuates between 0 and 12 for interaction of studied compounds with OASS (Fig. 14), and up to 15 for interaction of compounds 4 and 7 with PdxK (Fig. 15) throughout the trajectory. In addition, the various complexes showed several deep continuous interaction bands with many amino acid residues, connoting stable interactions between these ligands and several of the proteins’ residues in a fashion that compares favourably with the co-crystal. Therefore, the MD simulation study has provided insight into the favourability and stability of the protein-ligand interactions of the selected diselenides.