Screening of caspase-3 inhibitors from natural molecule database using e-pharmacophore and docking studies

Caspase a protease family member, have a vital role in cell death and inflammation process. Caspase-3, an effector caspase controls the regulation of apoptosis and has an anti apoptotic function. The mechanical significance of restoring apoptosis signaling to selectively target malignant cells is utilized to develop strong therapeutic strategies by the caspase family of mortality - induction molecules. Caspase-3 has currently no clear role in treatment for tumor progression and tumor sensitivity. The present study was aimed to screen caspase for potential inhibitors using computer aided docking methodologies. For this, zinc natural molecule database molecules were screened using e-pharmacophore and ADME protocols along with docking studies. Docking analysis selected two molecules, namely ZINC13341044 and ZINC13507846 with G-scores -5.27 and -6.19 respectively. These two potential hits are predicted as caspase inhibitors based on the results and can be further processed for in vitro validation.


Background:
The Extracellular matrix (ECM) receptors are imperative controllers of angiogenesis. One of these receptors, integrin α5β1, impact tumor-cell survival, multiplication, and metastasis, since the adversaries of this integrin α5β1 strongly restrain angiogenesis and tumor development [1]. Unligated α5β1 integrin inhibits survival and proliferation of the tumor cell even when they adhere to the ECM through different integrins assuming a major role in the direction of cell survival [2,3]. On the other hand, for certain biochemical and morphological changes during apoptosis, Caspase-3 is required. It is a frequently activated death protease, which cleaves a range of important cell proteins with numerous death signals. This is also important for cell death in a significant manner based on tissue, cell -type or death stimulus, as it is essential for the implementation and completion of apoptosis in certain types of characteristic cell morphology changes and biochemistry events. However, the specific requirements of this caspase in apoptosis were largely unknown [4]. Few reports show that integrin and caspases interact directly, although caspases were activated via integrin generated signaling pathways [5]. In the plasma membrane of the rat fibroblast cells during late stages of anoikis, our previous data reported the direct interaction between α5β1 integrin and caspase 3. These cells avoid cell death through the interaction of caspase 3 and unligated α5β1 integrins during the non -adherence process [2]. Screening of natural molecules for their biological activity using in vitro protocols is a time-consuming process and success ratio was also low. In silico methods became prominent in ©Biomedical Informatics (2019) screening of lead molecules by reducing experimental time and eliminating false positives. The aim of the present study was focused mainly on screening of small, potent inhibitors against caspase-3 protein.

Methodology: Protein preparation and grid generation:
The 3D crystalized structure of caspase protein (PDB ID: 5IBC) was retrieved from the protein data bank [6]. The protein was prepared by using the protein prep wizard [7], helps in converting the raw structure to a refined structure. Major steps in the preparation involve addition of hydrogen, removal of unwanted water molecules beyond 5Å, optimizing and minimizing the structure. The active pocket in the prepared protein was freezed by using the receptor grid generation.

Database preparation:
Zinc natural molecules database [8] were retrieved, conversion of molecules structure from 2D to 3D and refinement steps were carried using the canvas module from the Schrodinger software [9, 10]. Further through Conf-Gen application [11] the molecules confirmations were generated.

Pharmacophore hypothesis generation and database screening:
Two methodologies, structure-based drug-design and ligand-based drug-design are renowned important in silico screening approaches in drug discovery pipeline. e-Pharmacophore based methodology combines both structure-based and ligand-based methods to screen the molecule database [12,13]. Using the crystal protein-ligand complex a hypothesis was generated and further screened the database by setting the application values to default. Further the screened molecules were subjected to QikProp for ADME analysis [14].

Docking studies:
The screened molecules were docked into the active site of the caspase protein with the help of XP docking protocol [15-17] of the glide application. The application run was carried by choosing the protein grid file, screened molecules, setting docking protocol to XP and remaining options to default. The complexes were evaluated based on the binding modes between the protein and ligand along with the G-scores. The G-scores were calculated based on the following formula Glide score =0.065× vdW +0.130× Coul + Lipo + Hbond + Metal + BuryP + RotB + Site vdW -van der Waals energy, Coul -Coulomb energy, Lipo represents lipophilic term derived from hydrophobic grid potential, Hbond -hydrogen-bond, Metal -metal-binding term, BuryP -buried polar groups, RotB -penalty for freezing rotatable bonds, and Sitepolar interactions in the active site.

Results and Discussion:
Initially, for the generation of pharmacophore hypothesis the crystal protein was separated into protein and ligand. With the receptor grid generation application, a grid was generated around the active site and the crystal ligand was docked into that grid boxed active site using Glide XP protocol. The ligand orientation after docking was cross verified with crystal structure orientation, same orientation was reproduced confirming the docking protocol is valid and further used for docking studies.
©Biomedical Informatics (2019)  Screening studies: e-pharmacophore based screening was carried using the re-docked pose viewer file, a three site hypothesis RRD (R represents ring and D represents donor) was generated (figure 1). By setting all the parameters to default in the Phase screening protocol, the molecules database was screened with the hypothesis and a total of 32 hits were retrieved based on 3 out 3 matching criteria, i.e. all the three features of crystal molecule must be full filled by the screened molecule. Matched hits were further screened using QikProp application by choosing CNS, human oral absorption and Lipinski's rule of five as the major criteria's ( Table 1). Out of those 32 molecules, 7 molecules passed CNS as -2 (Predicted inactive against central nervous system), percent human oral absorption in between 80 to 100 and Lipinski's rule violations as zero (rule of five includes mol_MW < 500, QPlog Po/w < 5, donor HB ≤ 5, accpt HB ≤ 10 and the compounds which fulfill these rules were considered as druglike).

Docking studies:
The seven hits attained from the screening studies were subjected to molecular docking studies. Using Glide XP docking protocol, the molecules binding affinity with the amino acids present in the active pocket of the protein were studied.

Conclusion:
The main objective of our present study is to screen natural molecules database to select small molecules as inhibitors against caspase-3. Analysis selected two potential hits against the target namely ZINC13341044 and ZINC13507846 based on the binding mode with the target active site amino acids, satisfying all the ADME important descriptors and with good energy calculations. The caspase-3 inhibition activity for these molecules can be validated using in vitro and in vivo methods.

Conflict of Interest:
All authors have no conflict of interest