The process of pharmaceutical drug design usually begins the discovery of lead molecules and followed by its optimization which includes the synthesis and testing of thousands derivatives of lead structure in order to find the clinically useful drugs. In recent years many compounds have been discovered and implemented in treatment of Snake bites and designing of drug molecules (Table-2.1) was targeted on different types of Venoms with their function and activity in human body. In various categories a number of leading drug molecules have proven great success for the treatment of Snake bites from past to present. Traditional attempts for the treatment of Snake bites by using plants and herbs were not sufficient enough in many different types of snake bites. Antivenom containing antibodies are capable of holding the memory of particular type of snake venom and also capable to stop it's functioning at certain stage and site. Computational methodologies have grown have grown rapidly over the past few years and played an important role in the development of a number of drugs available in the market or undergoing clinical trials. Quantitative structure-activity relationship (QSAR) models have ability to predict the activity of new molecules. In this view efforts have been made understand three-dimensional quantitative structure activity relationships (3D-QSAR) with respect to enzyme phospholipase-A2. Previously developed CoMFA method provide significant value in terms of a new molecular design, when contours of the PLS coefficients are visualized for the set of molecules. Similarly, the kNN-MFA models provide direction for the design of new molecules in a rather convenient way. In the present study we reported 3D-QSAR studies on several Ursolic Acid derivatives and a correlation to predict the cardiotoxin venom activities degree of reliability using previously reported phospholipase-A2 inhibitors. The chemical structures of molecules and their experiment IC50 values are given in Table-3.1.
[4.2] kNN-MFA Analysis:
We developed a ligand based 3D QSAR model for predicting activity of ursolic acid like compounds using the kNN-MFA principle implemented on VlifeMDS software. K-nearest neighbor molecular field analysis (kNN-MFA) requires suitable alignment of given set of molecules. This is followed by automatically generated of a common rectangular grid around the molecules. The electrostatic and steric interaction energies are computed at the lattice points regularly spaced (2Å) grid using sp3 carbon probe of charge +1 and grid setting. Alignment of the molecules was done by template based method using the most active molecules as a reference molecules and substructure (Fig-3.1 & 3.2) define as a template structure for alignment of a set of molecules, shown in fig-3.3 & 3.4. In this study, we constructed several models for the given or selected member of training and test set of ursolic acid derivatives using random selection method and the best model with good statistical value are reported herein. We randomly selected 75%-65% (75%, 74%, 73%, 72%, 70%, 69%, 68% and 67%) of the total dataset compounds as a training set to build the model while the remaining 25%-33% (25%, 26%, 27%, 28%, 30%, 32% and 33%) of the total dataset compounds served as a test set to evaluate the robustness of the model and ten trials of each were run. After the selection of training and test dataset kNN-MFA was applied using stepwise (SW), simulated annealing (SA) and genetic algorithm (GA) approaches for building QSAR models. Among the all entire predictive models one good statistical model generate in stepwise forward-backward kNN-MFA method by random selection method having better statistical result is listed in table-3.4. Twenty two molecules (65% of total dataset compounds) out of the total thirty four phospholipase inhibitors were used as a training set to build the these predictive model while the remaining twelve compound (35% of total dataset compound) Table-3.5 marked with * served as a test set to evaluate the robustness of the model. The unicolumn statistics activity distribution plot observes that robustness of the model. The unicolumn statistics activity distribution plot observe that the activities of almost all the compounds in test set are lay within the range of minimum and maximum limit of the training set (Table-3.3 &3.4).
Result of the statistically best significant model having internal predictively or cross validated correlation coefficient 89% (= 0.8956), highest external predictively or predictive correlation coefficient 87% (pred_= 0.8756), Vn=2, k=3, degree of freedom F=19 and the steric field descriptors explain 100% of the variance, which support the statistical validity of the developed model, can be selected for predicted activities of the inhibitors vs their experimental activities of the training set and test set compounds listed in table-3.1 and the data fitness plot for significant model is shown in fig-3.9 which show the correlation between the predicted activities and the experimental values of the molecules. The plot of actual vs predicted activity of training and test sets for significant model is shown in fig-3.10 & 3.11 respectively. From the fitness plot it can see that kNN-MFa model is able to predict the activity of training set quite well (all points are close to regression line) as well external and that prediction by kNN-MFA model should have satisfactory predictive ability.
The kNN-MFA contour plot(fig-3.4) of generated models provide insight in to the various interactive fields on the activity and showed good statistical values with two significant steric descriptors S_712 (-0.3581, -0.2767) at the grid points with ranges of values shown in parenthesis represented in fig-3.4.The Negative range in steric descriptors indicates that negative steric potential is favorable to increase the activity of molecules and less bulky substituents or groups with low steric factor is preferred in that region, positive value of steric descriptors reveals that positive steric potential is favorable for increase in activity and more bulky substituents or groups with high steric factor is preferred in that region. Thus these kNN-MFA models provided further understanding of the relationship between structural features of ursolic acid derivatives and their activities which should be applicable to design novel potential drug candidates with highest potent activity in a rather convenient way.
[4.3] Validation of the 3D-QSAR model:
Significant 3D-QSAR model was produced by the randomly selected Twenty two molecules of the total dataset compounds as a training set to build the model and twelve compounds as a test set to verify the stability and predictive ability of the kNN-MFA model. The predicted IC50 value with the QSAR model are in good agreement with experimental values with a very well cross validation (= 0.8956) as well as external validation (pred_= 0.8756) of the model. The one steric descriptors S_712 (-0.3581, -0.2767) at the grid points shown in fig-3.4 play important role in designing new molecules with maximum potency. Negative range indicates that negative steric potential is favorable for increase in the activity and hence less bulky substituent group is preferred in that region. Compounds having more bulky substituent group is not favorable for biological activity in that region. Positive range indicates that positive steric potential is favorable for increase in the activity and hence more bulky substituent group is preferred in that region. Compounds having less bulky substituent group is not favorable for biological activity in that region. The testing result show that prediction by the kNN-MFA model is reasonably accurate and can be reliably used in the design of novel phospholipase inhibitors.
[4.4] Analysis of modified structure:
The analysis of the structure that has to be modifies is based on the positive result or signal shown by it towards the ligand. From 34 molecules that have aligned over the ligand, we divided them into the group of training sets and test sets. Now we have get total 22 training sets and remaining 12 test sets. Out of these 22 training sets we have got the output of 12 sets which shows the positive value (see table 3.5) during our process and that is the one thing we required to made changes into the structures and modified them.
[4.5] Molecular Docking:
Docking was performed using Autodock 4.2 (http://autodock.scripps.edu/). To verify the accuracy of docking results, the ligand ursolic acid was extracted from crystal structure in its experimental conformation and it was docked back to the corresponding binding pocket. The top ranking conformation clusters from this dock were evaluated in terms of root mean square deviation between docked position and experimentally determined position for the ligand. The low RMSD between the experimental and docked co-ordinates of ligand indicated energetically favorable and statistically validated docking result. Result of the control docking showed that Autodock4.2 determined the optimal orientation of the docked inhibitor, ursolic acid to be close to that of the original orientation found in the crystal structure.
Table-4.1: Modified Structures along with docking energies
omp. No.
Change Str. of derivative
Binding energy
KI
Intermolecular Energy (kcal/mol)
Internal Energy (kcal/mol)
Docking Energy (kcal/mol)
Torsional Energy (kcal/mol)
Ref.
R.M.S.
Comp-1
C:\Users\AMIT\Pictures\1.png
-6.91
8.67µM
-7.13
-0.05
-7.18
0.27
42.44
Comp-5
C:\Users\AMIT\Pictures\5.png
-6.7
12.22µM
-7.35
-0.18
-7.53
0.82
45.12
Comp-6
C:\Users\AMIT\Pictures\6.png
-5.0
217.9µM
-4.67
-5.54
-10.21
5.21
34.3
Comp-9
C:\Users\AMIT\Pictures\9.png
-3.63
2.19 mM
-5.16
-6.98
-12.14
8.51
38.86
Comp-18
C:\Users\AMIT\Pictures\18.png
-5.36
118.76µM
-7.23
-0.59
-7.82
2.47
40.31
Comp-19
C:\Users\AMIT\Pictures\19.png
-5.0
194.59µM
-6.3
0.14
-6.16
1.1
26.69
Comp-21
C:\Users\AMIT\Pictures\21.png
-5.73
62.59µM
-6.66
-0.17
-6.83
1.1
26.4
Comp-24
C:\Users\AMIT\Pictures\24.png
-6.21
28.10µM
-6.79
-0.24
-7.03
0.82
42.96
Comp-31
C:\Users\AMIT\Pictures\31.png
-6.54
15.97µM
-7.41
-0.24
-7.65
1.1
27.86
Comp-33
C:\Users\AMIT\Pictures\33.png
-6.18
10.22µM
-7.41
-0.22
-7.63
0.82
33.32
Comp-34
C:\Users\AMIT\Pictures\34.png
-6.53
16.22µM
-7.04
-0.04
-7.08
0.55
32.68
Comp-39
C:\Users\AMIT\Pictures\39.png
-5.76
60.16µM
-6.45
0.14
-6.31
0.55
46.59
[4.6] Binding interaction between ursolic acid derivatives with phospholipase-A2:
[4.6.1] Compound 1:
The compound 1 is seems to be bind perfectly over the ligand phospholipase A2 at the edge of its structure. The CYS77, TYR75, and LYS7 are the amino acids that looked to be take part in the binding reaction. We can signify them as the active sites of the ligand for the compound 1.
Fig-4.1: Compound 1
[4.6.2] Compound 5:
The compound 5 is seems to be bind perfectly over the ligand phospholipase A2 at the edge of its structure. The CYS77, TYR75, and LYS7 are again the same amino acids that looked to be take part in the binding reaction. We can signify them as the active sites of the ligand for the compound 5.
Fig-4.2: compound 5
[4.6.3] Compound 18:
The compound 18 is seems to be bind perfectly over the ligand phospholipase A2 at the edge of its structure. The PHE94, GLU90, ALA93, ASN97, and ASN101 are the amino acids that looked to be take part in the binding reaction. We can signify them as the active sites of the ligand for the compound 18.
C:\Users\AMIT\Desktop\amit new\New Folder\pro\docking-18\18-MS.png
Fig-4.3: compound 18
[4.6.4] Compound 31:
The compound 31 is seems to be bind perfectly over the ligand phospholipase A2 at the edge of its structure. The CYS77, TYR75, SER74 and SER76 are the amino acids that looked to be take part in the binding reaction. We can signify them as the active sites of the ligand for the compound 31.
Fig-4.4: compound 31
[4.6.5] Compound 33:
The compound 33 is seems to be bind perfectly over the ligand phospholipase A2 at the edge of its structure. The TYR75, THR83, SER76 and SER86 are the amino acids that looked to be take part in the binding reaction. We considered these amino acids as the active sites of the ligand for the compound 33.
Fig-4.5: compound 33
[4.6.6] Compound 34:
The compound 34 is seems to be bind perfectly over the ligand phospholipase A2 at the edge of its structure. The GLN4 and TYR75 are the only two amino acids that looked to be take part in the binding reaction. They are the only active sites of the ligand for the compound 34.
Fig-4.6: compound 34
