Discovery of Novel 1,2,4-Oxadiazole Derivatives as Potent Caspase-3 Activator for Cancer Treatment

Abstract

In the present study, a quantitative structure–activity relationship (QSAR) and docking studies were accomplished on a series of 1,2,4-oxadiazoles. The results of QSARs are reliable and have high predictive ability for both the internal (q² = 0.610) and external (pred_r² = 0.553) datasets with least standard error (SE; i.e., 0.130) and four principal components, which signifies the reliability of the generated model. Molecular docking was also reported by the GOLD docking program, which showed that the hydrogen bonding may be responsible for the activity, and may be further increased upon adding high electronegative substitutions.

1. Introduction

Cancer is the second leading cause of mortality globally and becomes the principal cause of mortality in developing countries [1,2,3]. Recently, several anticancer drugs have become available on the market, acting through different mechanisms; however, the majority of them are associated with serious side effects [4,5]. The lack of targeting ability of these drugs is responsible for their side effects [6]. Large numbers of heterocyclic compounds have been reported for potential anticancer activities [7]. These heterocyclic-based anticancer agents are either under investigation or marketed as potent anticancer agents [8,9,10,11]. Oxadiazole is an important five-membered heterocyclic compound, having one oxygen atom and two nitrogen atoms. Recently, several 1,2,4-oxadiazole derivatives have been shown to possess anticancer activity [12,13].

Nowadays, researchers are focusing on various groups of molecules that are involved in the apoptosis inducing cytotoxicity. Caspases, a group of cysteine proteases, are the executioners of apoptosis. These caspases cleave their substrates after aspartic acid residues. Initiator caspases (caspase 2/8/9/10) and effector caspases (caspase 3/6/7) are the two classes of caspase. Recently, the activation of caspase-3 mediated apoptosis becomes an interesting therapeutic strategy for cancer therapy. Zhang et al. reported a series of 3-Aryl-5-aryl-1,2,4-oxadiazoles as a novel apoptosis inducer through caspase-3 activation. The compounds’ activities have been reported against breast and colorectal cancer cell lines [14].

The quantitative structure–activity relationship (QSARs) is an attempt to correlate the structural features of the compounds quantitatively with their biological activities. Researchers reported thousands of QSAR studies in the search for novel anticancer agents [15,16,17,18,19].

In the search for new anticancer agents, our research group previously reported QSAR studies of 1,2,4-oxadiazole derivatives describing the key structure features responsible for anticancer activities [20,21,22]. In the continuation of our previous work, herein we report the two-dimensional QSAR (2D-QSAR) and molecular docking studies’ outcomes.

The 2D-QSAR studies were done using Step Wise k Nearest Neighbor Molecular Field Analysis [(SW) kNN MFA] using V-Life Molecular Design Software Version 3.0 (V-Life Molecular Design). The docking studies were also performed using GOLD software.

2. Material and Methods

2.1. Dataset

A dataset of twenty eight 3-aryl-5-aryl-1,2,4-oxadiazoles derivatives has been taken for present QSAR study (

Table 1. 1,2,4-Oxadiazole analogues and their experimental caspase-3 activator activity.

S. No.	Compound	Ar1	Ar2	Experimental Activity pEC50 (nM) (DLD1)
1	1d			3.357
2	4a			3.102
3	4b			3.119
4	4c			3.367
5	4d			2.839
6	4e			2.848
7	4g			3.553
8	4h			3.432
9	4i			3.252
10	4j			3.420
11	4k			3.387
12	4l			3.409
13	4m			3.620
14	4n			3.538
15	4o			2.879
16	10a			3.081
17	10b			2.827
18	10d			2.971
19	10e			3.319
20	10f			3.076
21	10g			3.155
22	10h			3.237
23	11a			3.229
24	11b			3.236
25	11c			2.959
26	11d			2.959
27	11e			3.149
28	11f			2.921

). Compounds have high structural diversity with ample range of biological activity [14,23].

2.2. 2D QSAR

2D QSAR studies were performed via Step Wise k Nearest Neighbor Molecular Field Analysis [(SW) kNN MFA] method using V-Life Molecular Design Software Version 3.0 (V-Life Molecular Design) [24,25,26].

The 2D QSAR studies were performed by dividing compounds in the training and test dataset which resulted several QSAR equations. Unicolumn statistics was done to divide training and test data compounds. Twenty-two compounds were positioned in the training set and 6 compounds (4b, 4d, 4e, 10a, 10d and 11h) in the test set.

2.3. Molecular Docking Analysis

Molecular docking was employed to locate the appropriate binding orientations and conformations of these 1,2,4-oxadiazoles interacting with caspase-3 using the docking program GOLD version 3.2. Ten docked conformers were produced for each 1,2,4-oxadiazole derivative. The conformation with the lowest docking energy in the most populated cluster is selected as the possible “active” conformation against the 1RE1 active site. In the present study, 28 compounds were successfully docked into the 1RE1 site.

The X-ray crystal structure (pdb: 1RE1) of caspase-3 was obtained from the Protein Data Bank. Initially, for protein preparation, water molecules were removed, hydrogen atoms added and AMBER7FF99 charges to the protein were applied. The ligands were docked inside a cubic GRID box (within 5 A° surrounding to the cocrystallized ligand) centered at the midpoint between the Cys205 and Gly238. Ten docking runs were performed for each compound in the dataset. In most cases the chosen pose was the top ranked solution.

3. Results and Discussion

3.1. 2D QSAR Results

The results of the unicolumn statistics are summarized in

Table 2. Unicolumn statistical data of training and test set in 2D quantitative structure–activity relationship (QSAR) models.

	Average	Maxima	Minima	Std. Deviation
Training set	3.174	3.620	2.827	0.228
Test set	3.234	3.553	2.848	0.264

, which showed that the test is interpolative i.e., both test and training dataset contain compounds of high structure diversity with variation in biological activity. The test and the training set contained a diverse set of compounds with low, moderate and high biological activity.

Finally, the following model was selected.

pEC₅₀ = 0.243 * IP − 0.139 * BC + 0.155 * DM + 0.008 * PSA + 0.0005

(1)

The obtained model showed a high correlation coefficient (r = 0.862) between descriptors including ionization potential (IP), bromine count (BC), dipole moment (DM), polar surface area (PSA) and anticancer activities. The squared correlation coefficient (r²) of 0.743, explains 74.29% of the variance in biological activity. The obtained model is statistical significant with F values F(4,21) = 11.561. The obtained model showed both good internal and external predictive ability with cross-validated squared correlation coefficient for internal dataset (q²) value 0.610 and for external dataset (pred_r²) value 0.553 with a standard error (SE) of 0.130 (

Table 3. Parameters value for the best 2D QSAR model generated.

Model	r	r2	q2	SE (r2 se)	Pred_r2	F-Value	Descriptors
1	0.862	0.743	0.610	0.130	0.553	11.561	IP, BC, DM, PSA

).

In the model, the contribution of the descriptors is presented in the contribution chart (Figure 1), signifying the positive contribution of the ionization potential (IP), dipole moment (DM) and polar surface area (PSA) towards the biological activity. The addition of substitution that increases the polarity of the compounds results in increased anticancer activity. The negative contribution of the bromine count signifies the lower number of bromine encouraging biological activities.

The correlation between the experimental and predicted activity of the compounds is shown in

Table 4. Experimental, predicted and residual activities of the compounds obtained in 2D QSAR and GOLD score.

		2D QSAR		Docking
Comp. No.	Experimental pEC50	[(SW) kNN MFA] Predicted pEC50	[(SW) kNN MFA] Residual	GOLD Docking
1d	3.357	3.291	0.065	50.771
4a	3.102	3.155	−0.052	48.324
4b	3.119	3.106	0.013	47.105
4c	3.367	3.300	0.067	51.672
4d	2.839	2.912	−0.074	52.859
4e	2.848	3.102	−0.255	50.846
4g	3.553	3.269	0.284	50.357
4h	3.432	3.306	0.126	50.212
4i	3.252	3.382	−0.130	50.304
4j	3.420	3.106	0.315	56.319
4k	3.387	3.415	−0.028	50.294
4l	3.409	3.361	0.048	51.432
4m	3.620	3.620	−0.0002	49.303
4n	3.538	3.532	0.005	50.930
4o	2.879	3.025	−0.146	50.968
10a	3.081	2.899	0.182	49.383
10b	2.827	3.045	−0.218	49.680
10d	2.971	3.128	−0.157	50.205
10e	3.319	3.233	0.086	51.265
10f	3.076	3.162	−0.086	49.302
10g	3.155	3.253	−0.098	49.203
10h	3.237	3.295	−0.058	49.839
11a	3.229	3.102	0.127	50.212
11b	3.236	3.234	0.003	49.423
11c	2.959	3.007	−0.049	45.962
11d	2.959	3.007	−0.049	47.860
11e	3.149	2.927	0.222	49.377
11f	2.921	2.879	0.042	50.891

and represented in Figure 2.

3.2. GOLD Docking Studies

All 28, 1,2,4-oxadiazoles derivatives were docked into the binding site of caspase-3 and the energy scores of the activators are also shown in Table 4. A precise correlation was observed in between docking scores and pIC₅₀ values.

A complete overview of GOLD docking is presented in Figure 3 and Figure 4.

The docking results revealed that most active compound 4m is properly located at the binding site of the Cys205 and Gly238 amino acid residues and numerous interactions occur between it and the binding region of the enzyme. The four key hydrogen bond interactions occur: (1) between the NH of Gly238 and the O of the oxadiazole ring; (2) between the NH of Gly238 and the N of the oxadiazole ring; (3) between the NH of Cys285 and the N of the oxadiazole ring; (4) between the NH of THR288 and the N of the pyridine ring residue (Figure 3). The hydrogen bonding distances observed were 1.549 Å (O···H-NH-Gly238), 2.712 Å (N···H-NH-Gly238), 2.429 Å (N···H-NH-Cys285) and 2.092 Å between the N of the pyridine ring and NH of THR288 (N···H-NH-THR288).

Akin to compound 4m, compound 10b was also docked at the same binding pockets having Cys205 and Gly238 amino acid residues (Figure 4). The result showed the formation of two hydrogen bonds: (1) between the NH of Cys205 and the O of the oxadiazole ring (O···H-NH-Cys205), having 2.145 Å bond distance; (2) between the NH of Cys205 and the N of the oxadiazole ring (N···H-NH-Cys205) with 2.614 Å bond length.

The docking results revealed that the hydrogen bonding may be responsible for biological activity, which may be further increase upon adding more electronegative substitutions. The correlation between the dock score and the experimental activity is shown graphically in Figure 5, which shows a linear correlation between the dock score and biological activity.

The results of the QSAR analysis clearly show that upon increasing the polarity in terms of theionization potential (IP), dipole moment (DM) and polar surface area (PSA), biological activity will also be enhanced. The docking results also support the QSAR outcomes.

4. Conclusions

In conclusion, the current QSAR studies established a reliable QSAR model with high predictive ability with q² = 0.610, r² = 0.743 and low standard error (SE) = 0.130 and four principal components. The predicted value of the external test set (pred_r²) was also high (i.e., 0.553). The developed model was reliable, which indicated the importance of substitution in 1,2,4-oxadiazoles at their respective positions to improve anticancer activity. The positive contribution of ionization potential (IP), dipole moment (DM) and polar surface area (PSA) is conducive for biological activity, and further addition of these substitutions increases anticancer activity, while the negative contribution of the bromine count signifies the lower number of bromine encouraging the biological activities. The docking results explore the binding mode between the ligands and the receptor.